This commit is contained in:
2026-03-27 00:01:54 -03:00
parent 65814b5b9e
commit df6bcb01e8
14 changed files with 1246 additions and 203 deletions

View File

@@ -116,6 +116,65 @@ def list_checkpoints(job_id: str) -> list[CheckpointInfo]:
return result return result
class CheckpointFrameInfo(BaseModel):
seq: int
timestamp: float
jpeg_b64: str
class CheckpointData(BaseModel):
job_id: str
stage: str
profile_name: str
video_path: str
is_scenario: bool
scenario_label: str
frames: list[CheckpointFrameInfo]
stats: dict = {}
config_snapshot: dict = {}
stage_output_key: str = ""
@router.get("/checkpoints/{job_id}/{stage}", response_model=CheckpointData)
def get_checkpoint_data(job_id: str, stage: str):
"""Load checkpoint frames + metadata for the editor UI."""
from core.db.detect import get_stage_checkpoint
from detect.checkpoint.frames import load_frames_b64
checkpoint = get_stage_checkpoint(job_id, stage)
if not checkpoint:
raise HTTPException(status_code=404, detail=f"No checkpoint for {job_id}/{stage}")
raw_manifest = checkpoint.frames_manifest or {}
manifest = {int(k): v for k, v in raw_manifest.items()}
frame_metadata = checkpoint.frames_meta or []
# Only load filtered frames if available, otherwise all
filtered = set(checkpoint.filtered_frame_sequences or [])
if filtered:
manifest = {k: v for k, v in manifest.items() if k in filtered}
frames_b64 = load_frames_b64(manifest, frame_metadata)
frame_list = [
CheckpointFrameInfo(seq=f["seq"], timestamp=f["timestamp"], jpeg_b64=f["jpeg_b64"])
for f in frames_b64
]
return CheckpointData(
job_id=str(checkpoint.job_id),
stage=checkpoint.stage,
profile_name=checkpoint.profile_name,
video_path=checkpoint.video_path,
is_scenario=checkpoint.is_scenario,
scenario_label=checkpoint.scenario_label,
frames=frame_list,
stats=checkpoint.stats or {},
config_snapshot=checkpoint.config_snapshot or {},
stage_output_key=checkpoint.stage_output_key or "",
)
@router.get("/scenarios", response_model=list[ScenarioInfo]) @router.get("/scenarios", response_model=list[ScenarioInfo])
def list_scenarios_endpoint(): def list_scenarios_endpoint():
"""List all available scenarios (bookmarked checkpoints).""" """List all available scenarios (bookmarked checkpoints)."""

View File

@@ -110,7 +110,7 @@ class StageCheckpoint:
# Scenario — a checkpoint bookmarked for the editor workflow. # Scenario — a checkpoint bookmarked for the editor workflow.
# Created by seeders (manual scripts that populate state from real footage) # Created by seeders (manual scripts that populate state from real footage)
# or captured from a running pipeline. Loaded via URL: # or captured from a running pipeline. Loaded via URL:
# /detection/?job=<job_id>&stage=<stage>&editor=true # /detection/?job=<job_id>#/editor/<stage>
is_scenario: bool = False is_scenario: bool = False
scenario_label: str = "" # human-readable name, e.g. "chelsea_edges_lowcanny" scenario_label: str = "" # human-readable name, e.g. "chelsea_edges_lowcanny"

View File

@@ -78,3 +78,34 @@ def load_frames(manifest: dict[int, str], frame_metadata: list[dict]) -> list[Fr
frames.sort(key=lambda f: f.sequence) frames.sort(key=lambda f: f.sequence)
return frames return frames
def load_frames_b64(manifest: dict[int, str], frame_metadata: list[dict]) -> list[dict]:
"""
Load frame images from S3 as base64 JPEG — lightweight, no numpy.
Returns list of dicts: {seq, timestamp, jpeg_b64}
"""
import base64
from core.storage.s3 import download_to_temp
meta_map = {m["sequence"]: m for m in frame_metadata}
frames = []
for seq, key in manifest.items():
tmp_path = download_to_temp(BUCKET, key)
try:
with open(tmp_path, "rb") as f:
jpeg_bytes = f.read()
finally:
os.unlink(tmp_path)
meta = meta_map.get(seq, {})
frames.append({
"seq": seq,
"timestamp": meta.get("timestamp", 0.0),
"jpeg_b64": base64.b64encode(jpeg_bytes).decode(),
})
frames.sort(key=lambda f: f["seq"])
return frames

View File

@@ -20,7 +20,7 @@ Usage:
python tests/detect/manual/seed_scenario.py --video media/mpr/out/chunks/.../chunk_0001.mp4 python tests/detect/manual/seed_scenario.py --video media/mpr/out/chunks/.../chunk_0001.mp4
Then open: Then open:
http://mpr.local.ar/detection/?job=<JOB_ID>&stage=filter_scenes&editor=true http://mpr.local.ar/detection/?job=<JOB_ID>#/editor/detect_edges
""" """
from __future__ import annotations from __future__ import annotations

View File

@@ -78,6 +78,18 @@ function onTimelineResize(delta: number) {
tableFlex.value = Math.max(0.3, Math.min(3, tableFlex.value - shift)) tableFlex.value = Math.max(0.3, Math.min(3, tableFlex.value - shift))
} }
// Editor sliders sidebar width — drag right = shrink sliders (grow frame)
const slidersWidth = ref(210)
function onSlidersResize(delta: number) {
slidersWidth.value = Math.max(210, Math.min(350, slidersWidth.value - delta))
}
// Editor bottom height (overlays bar)
const editorBottomHeight = ref(50)
function onEditorBottomResize(delta: number) {
editorBottomHeight.value = Math.max(36, Math.min(120, editorBottomHeight.value - delta))
}
const statusMap: Record<string, 'idle' | 'live' | 'processing' | 'error'> = { const statusMap: Record<string, 'idle' | 'live' | 'processing' | 'error'> = {
idle: 'idle', idle: 'idle',
connecting: 'processing', connecting: 'processing',
@@ -110,32 +122,108 @@ async function stopPipeline() {
const currentFrameImage = ref<string | null>(null) const currentFrameImage = ref<string | null>(null)
const currentFrameRef = ref<number | null>(null) const currentFrameRef = ref<number | null>(null)
// All checkpoint frames (for scenario mode — scrubbing)
const checkpointFrames = ref<{ seq: number; timestamp: number; jpeg_b64: string }[]>([])
const checkpointFrameIndex = ref(0)
const checkpointStage = ref<string | null>(null) // which stage the checkpoint is at
source.on<{ frame_ref: number; jpeg_b64: string }>('frame_update', (e) => { source.on<{ frame_ref: number; jpeg_b64: string }>('frame_update', (e) => {
currentFrameImage.value = e.jpeg_b64 currentFrameImage.value = e.jpeg_b64
currentFrameRef.value = e.frame_ref currentFrameRef.value = e.frame_ref
}) })
// Load checkpoint data when in scenario mode
async function loadCheckpoint(job: string, stage: string) {
try {
const resp = await fetch(`/api/detect/checkpoints/${job}/${stage}`)
if (!resp.ok) return
const data = await resp.json()
checkpointFrames.value = data.frames ?? []
checkpointStage.value = stage
// Show first frame
if (checkpointFrames.value.length > 0) {
checkpointFrameIndex.value = 0
const first = checkpointFrames.value[0]
currentFrameImage.value = first.jpeg_b64
currentFrameRef.value = first.seq
}
status.value = 'idle'
} catch (e) {
console.error('Failed to load checkpoint:', e)
}
}
function setCheckpointFrame(index: number) {
if (index < 0 || index >= checkpointFrames.value.length) return
checkpointFrameIndex.value = index
const frame = checkpointFrames.value[index]
currentFrameImage.value = frame.jpeg_b64
currentFrameRef.value = frame.seq
}
// Load checkpoint when in editor mode with a job (scenario URL)
// Uses watch to handle both initial load and navigation
import { watch as vueWatch } from 'vue'
vueWatch(
() => [pipeline.layoutMode, pipeline.editorStage, jobId.value] as const,
([mode, stage, job]) => {
if (mode === 'bbox_editor' && stage && job) {
const stageMap: Record<string, string> = {
detect_edges: 'filter_scenes',
detect_contours: 'detect_edges',
detect_color: 'detect_contours',
merge_regions: 'detect_color',
}
const cpStage = stageMap[stage] ?? 'filter_scenes'
loadCheckpoint(job, cpStage)
}
},
{ immediate: true },
)
// Debug overlays from replay-stage results // Debug overlays from replay-stage results
const editorOverlays = ref<FrameOverlay[]>([]) const editorOverlays = ref<FrameOverlay[]>([])
// Boxes from edge detection (local or server)
const editorBoxes = ref<import('mpr-ui-framework/src/renderers/FrameRenderer.vue').FrameBBox[]>([])
function onReplayResult(result: { function onReplayResult(result: {
regions_by_frame?: Record<string, unknown[]>
debug?: Record<string, { edge_overlay_b64: string; lines_overlay_b64: string; horizontal_count: number; pair_count: number }> debug?: Record<string, { edge_overlay_b64: string; lines_overlay_b64: string; horizontal_count: number; pair_count: number }>
}) { }) {
const overlays: FrameOverlay[] = [] // Update boxes
if (result.regions_by_frame) {
const firstRegions = Object.values(result.regions_by_frame)[0] as any[] ?? []
editorBoxes.value = firstRegions.map((r: any) => ({
x: r.x, y: r.y, w: r.w, h: r.h,
confidence: r.confidence,
label: r.label ?? 'edge_region',
stage: 'detect_edges',
}))
}
// Update overlays — only when debug data is present, preserve existing otherwise
if (result.debug) { if (result.debug) {
// Take first frame's debug data (editor shows one frame at a time)
const firstDebug = Object.values(result.debug)[0] const firstDebug = Object.values(result.debug)[0]
if (firstDebug) { if (firstDebug) {
const overlays: FrameOverlay[] = []
if (firstDebug.edge_overlay_b64) { if (firstDebug.edge_overlay_b64) {
overlays.push({ src: firstDebug.edge_overlay_b64, label: 'Canny edges', visible: true, opacity: 0.4 }) // Preserve visibility/opacity from existing overlays if they exist
const existing = editorOverlays.value.find(o => o.label === 'Canny edges')
overlays.push({ src: firstDebug.edge_overlay_b64, label: 'Canny edges', visible: existing?.visible ?? true, opacity: existing?.opacity ?? 0.25 })
} }
if (firstDebug.lines_overlay_b64) { if (firstDebug.lines_overlay_b64) {
overlays.push({ src: firstDebug.lines_overlay_b64, label: 'Hough lines', visible: true, opacity: 0.5 }) const existing = editorOverlays.value.find(o => o.label === 'Hough lines')
} overlays.push({ src: firstDebug.lines_overlay_b64, label: 'Hough lines', visible: existing?.visible ?? true, opacity: existing?.opacity ?? 0.25 })
}
} }
editorOverlays.value = overlays editorOverlays.value = overlays
} }
}
}
function onJobStarted(newJobId: string) { function onJobStarted(newJobId: string) {
jobId.value = newJobId jobId.value = newJobId
@@ -236,12 +324,14 @@ function onJobStarted(newJobId: string) {
<!-- === BBOX EDITOR MODE === --> <!-- === BBOX EDITOR MODE === -->
<template v-else-if="pipeline.layoutMode === 'bbox_editor'"> <template v-else-if="pipeline.layoutMode === 'bbox_editor'">
<Panel :title="`Region Editor — ${pipeline.editorStage?.replace(/_/g, ' ')}`" :status="status"> <div class="editor-layout">
<div class="editor-placeholder"> <!-- Top: frame + sliders side by side -->
<div class="editor-top">
<div class="editor-frame"> <div class="editor-frame">
<FramePanel :source="source" :status="status" :overlays="editorOverlays" /> <FramePanel :source="source" :status="status" :overlays="editorOverlays" :frame-image="currentFrameImage" :editor-boxes="editorBoxes" />
</div> </div>
<div class="editor-tools"> <ResizeHandle direction="horizontal" @resize="onSlidersResize" />
<div class="editor-sliders" :style="{ width: slidersWidth + 'px' }">
<StageConfigSliders <StageConfigSliders
v-if="pipeline.editorStage" v-if="pipeline.editorStage"
:stage="pipeline.editorStage" :stage="pipeline.editorStage"
@@ -250,10 +340,29 @@ function onJobStarted(newJobId: string) {
:frame-ref="currentFrameRef" :frame-ref="currentFrameRef"
@replay-result="onReplayResult" @replay-result="onReplayResult"
/> />
</div>
</div>
<!-- Bottom: debug overlays + close -->
<div class="editor-bottom">
<div class="overlay-controls">
<template v-if="editorOverlays.length > 0">
<label v-for="(overlay, idx) in editorOverlays" :key="idx" class="overlay-toggle">
<input type="checkbox" v-model="overlay.visible" />
<span class="overlay-label">{{ overlay.label }}</span>
<input
type="range"
min="0" max="1" step="0.05"
:value="overlay.opacity ?? 0.5"
@input="(e: Event) => overlay.opacity = Number((e.target as HTMLInputElement).value)"
class="opacity-slider"
/>
<span class="opacity-value">{{ Math.round((overlay.opacity ?? 0.5) * 100) }}%</span>
</label>
</template>
</div>
<button class="editor-close" @click="pipeline.closeEditor()">✕ Close</button> <button class="editor-close" @click="pipeline.closeEditor()">✕ Close</button>
</div> </div>
</div> </div>
</Panel>
</template> </template>
<!-- === STAGE EDITOR MODE === --> <!-- === STAGE EDITOR MODE === -->
@@ -279,28 +388,8 @@ function onJobStarted(newJobId: string) {
<!-- Bottom bar: Log or Blob viewer depending on mode --> <!-- Bottom bar: Log or Blob viewer depending on mode -->
<div class="log-row"> <div class="log-row">
<template v-if="pipeline.layoutMode === 'bbox_editor'"> <template v-if="pipeline.layoutMode === 'source_selector'">
<Panel :title="`Debug Overlays — ${pipeline.editorStage?.replace(/_/g, ' ')}`" :status="status"> <!-- no log in source selector -->
<div class="overlay-controls">
<template v-if="editorOverlays.length > 0">
<label v-for="(overlay, idx) in editorOverlays" :key="idx" class="overlay-toggle">
<input type="checkbox" v-model="overlay.visible" />
<span class="overlay-label">{{ overlay.label }}</span>
<input
type="range"
min="0" max="1" step="0.05"
:value="overlay.opacity ?? 0.5"
@input="(e) => overlay.opacity = Number((e.target as HTMLInputElement).value)"
class="opacity-slider"
/>
<span class="opacity-value">{{ Math.round((overlay.opacity ?? 0.5) * 100) }}%</span>
</label>
</template>
<div v-else class="blob-placeholder">
Run analysis with debug enabled to see edge and line overlays
</div>
</div>
</Panel>
</template> </template>
<template v-else> <template v-else>
<LogPanel ref="logPanel" :source="source" :status="status" /> <LogPanel ref="logPanel" :source="source" :status="status" />
@@ -478,34 +567,82 @@ header h1 { font-size: var(--font-size-lg); font-weight: 600; }
/* Log: full width bottom */ /* Log: full width bottom */
.log-row { .log-row {
flex-shrink: 0; flex-shrink: 0;
height: 200px; height: 150px;
} }
.empty { color: var(--text-dim); padding: var(--space-6); text-align: center; } .empty { color: var(--text-dim); padding: var(--space-6); text-align: center; }
/* Editor placeholders */ /* Editor layout — frame maximized, sliders right, overlays bottom */
.editor-placeholder { .editor-layout {
display: flex; display: flex;
flex-direction: column;
height: 100%; height: 100%;
gap: var(--space-2); min-height: 0;
overflow: hidden;
}
.editor-top {
display: flex;
flex: 1;
min-height: 0;
overflow: hidden;
} }
.editor-frame { .editor-frame {
flex: 1; flex: 1;
min-height: 0; min-height: 0;
min-width: 0;
overflow: hidden;
display: flex;
}
.editor-frame > * {
flex: 1;
min-height: 0;
overflow: hidden;
} }
.editor-tools { .editor-sliders {
width: 200px;
flex-shrink: 0; flex-shrink: 0;
padding: var(--space-3); min-width: 210px;
padding: var(--space-2);
background: var(--surface-2); background: var(--surface-2);
border-radius: var(--panel-radius); overflow-y: auto;
overflow-x: hidden;
}
.editor-bottom {
flex-shrink: 0;
display: flex; display: flex;
flex-direction: column; align-items: center;
gap: var(--space-2); gap: var(--space-4);
font-size: var(--font-size-sm); padding: var(--space-2) var(--space-3);
background: var(--surface-2);
border-top: var(--panel-border);
height: 36px;
}
.editor-close {
background: var(--surface-3);
border: 1px solid var(--surface-3);
border-radius: 4px;
padding: 3px 10px;
color: var(--text-secondary); color: var(--text-secondary);
font-family: var(--font-mono);
font-size: var(--font-size-sm);
cursor: pointer;
margin-left: auto;
}
.editor-close:hover {
background: var(--status-error);
color: #000;
}
/* Stage config editor (placeholder) */
.editor-placeholder {
display: flex;
height: 100%;
gap: var(--space-2);
} }
.editor-config { .editor-config {
@@ -517,23 +654,6 @@ header h1 { font-size: var(--font-size-lg); font-weight: 600; }
gap: var(--space-2); gap: var(--space-2);
} }
.editor-close {
background: var(--surface-3);
border: 1px solid var(--surface-3);
border-radius: 4px;
padding: var(--space-2) var(--space-3);
color: var(--text-secondary);
font-family: var(--font-mono);
font-size: var(--font-size-sm);
cursor: pointer;
margin-top: auto;
}
.editor-close:hover {
background: var(--status-error);
color: #000;
}
.blob-placeholder { .blob-placeholder {
padding: var(--space-4); padding: var(--space-4);
color: var(--text-dim); color: var(--text-dim);

View File

@@ -1,9 +1,16 @@
<script setup lang="ts"> <script setup lang="ts">
import { ref, onMounted, computed } from 'vue' import { ref, onMounted, computed, watch } from 'vue'
import {
runEdgeDetection,
runEdgeDetectionDebug,
b64ToImageData,
imageDataToB64,
type EdgeDetectionParams,
} from 'mpr-ui-framework/src/cv'
interface ConfigField { interface ConfigField {
name: string name: string
type: string // "bool" | "int" | "float" | "str" type: string
default: unknown default: unknown
description: string description: string
min: number | null min: number | null
@@ -14,16 +21,15 @@ interface ConfigField {
const props = defineProps<{ const props = defineProps<{
/** Stage name (e.g. "detect_edges") */ /** Stage name (e.g. "detect_edges") */
stage: string stage: string
/** Job ID for replay-stage calls (used as fallback) */ /** Job ID (used for server mode fallback) */
jobId: string jobId: string
/** Currently displayed frame image (base64 JPEG) — sent directly to GPU for fast feedback */ /** Currently displayed frame image (base64 JPEG) */
frameImage?: string | null frameImage?: string | null
/** Currently displayed frame sequence number */ /** Currently displayed frame sequence number */
frameRef?: number | null frameRef?: number | null
}>() }>()
const emit = defineEmits<{ const emit = defineEmits<{
/** Emitted when replay returns new regions */
'replay-result': [result: { 'replay-result': [result: {
regions_by_frame: Record<string, unknown[]> regions_by_frame: Record<string, unknown[]>
debug: Record<string, { edge_overlay_b64: string; lines_overlay_b64: string; horizontal_count: number; pair_count: number }> debug: Record<string, { edge_overlay_b64: string; lines_overlay_b64: string; horizontal_count: number; pair_count: number }>
@@ -36,24 +42,50 @@ const loading = ref(false)
const error = ref<string | null>(null) const error = ref<string | null>(null)
const regionCount = ref<number | null>(null) const regionCount = ref<number | null>(null)
const debugEnabled = ref(true) const debugEnabled = ref(true)
const autoApply = ref(true) // auto-run on slider change (fast CV); uncheck for heavy stages
const execMode = ref<'local' | 'server'>('local')
const execTimeMs = ref<number | null>(null)
// Config field defaults for detect_edges (used when API is unavailable)
const EDGE_DEFAULTS: ConfigField[] = [
{ name: 'enabled', type: 'bool', default: true, description: 'Enable edge detection', min: null, max: null, options: null },
{ name: 'edge_canny_low', type: 'int', default: 50, description: 'Canny low threshold', min: 0, max: 255, options: null },
{ name: 'edge_canny_high', type: 'int', default: 150, description: 'Canny high threshold', min: 0, max: 255, options: null },
{ name: 'edge_hough_threshold', type: 'int', default: 80, description: 'Hough accumulator threshold', min: 1, max: 500, options: null },
{ name: 'edge_hough_min_length', type: 'int', default: 100, description: 'Min line length (px)', min: 10, max: 2000, options: null },
{ name: 'edge_hough_max_gap', type: 'int', default: 10, description: 'Max line gap (px)', min: 1, max: 100, options: null },
{ name: 'edge_pair_max_distance', type: 'int', default: 200, description: 'Max pair distance (px)', min: 10, max: 500, options: null },
{ name: 'edge_pair_min_distance', type: 'int', default: 15, description: 'Min pair distance (px)', min: 5, max: 200, options: null },
]
// Fetch stage config fields from API
onMounted(async () => { onMounted(async () => {
// Try loading from API, fall back to hardcoded defaults
try { try {
const resp = await fetch(`/api/detect/config/stages/${props.stage}`) const resp = await fetch(`/api/detect/config/stages/${props.stage}`)
if (!resp.ok) { if (resp.ok) {
error.value = `Failed to load config: ${resp.status}`
return
}
const data = await resp.json() const data = await resp.json()
fields.value = data.config_fields ?? [] fields.value = data.config_fields ?? []
} else {
fields.value = EDGE_DEFAULTS
}
} catch {
fields.value = EDGE_DEFAULTS
}
// Initialize values from defaults
for (const f of fields.value) { for (const f of fields.value) {
values.value[f.name] = f.default values.value[f.name] = f.default
} }
} catch (e) {
error.value = `Failed to load config: ${e}` // Auto-run on first frame if already available
if (props.frameImage) {
applyDetection()
}
})
// Auto-run when frame arrives after mount (checkpoint load is async)
watch(() => props.frameImage, (newVal, oldVal) => {
if (newVal && !oldVal && fields.value.length > 0) {
applyDetection()
} }
}) })
@@ -64,41 +96,97 @@ function resetDefaults() {
for (const f of fields.value) { for (const f of fields.value) {
values.value[f.name] = f.default values.value[f.name] = f.default
} }
applyDetection()
} }
let debounceTimer: ReturnType<typeof setTimeout> | null = null let debounceTimer: ReturnType<typeof setTimeout> | null = null
function onSliderChange() { function onSliderChange() {
// Debounce — wait 300ms after last change before calling replay if (!autoApply.value) return
if (debounceTimer) clearTimeout(debounceTimer) if (debounceTimer) clearTimeout(debounceTimer)
debounceTimer = setTimeout(() => applyReplay(), 300) debounceTimer = setTimeout(() => applyDetection(), 150)
}
async function applyDetection() {
if (!props.frameImage) {
error.value = 'No frame available'
return
} }
async function applyReplay() {
loading.value = true loading.value = true
error.value = null error.value = null
execTimeMs.value = null
// Direct GPU call — send the frame image + current slider params try {
// Skip checkpoint/replay path for ~50-100ms round trips instead of seconds if (execMode.value === 'local') {
if (props.frameImage && props.stage === 'detect_edges') { await runLocal()
await callGpuDirect() } else {
return await runServer()
} }
} catch (e) {
// Fallback: replay-stage path (for stages without direct GPU endpoint) error.value = `${execMode.value} failed: ${e}`
if (!props.jobId) { } finally {
error.value = 'No frame image or job ID available'
loading.value = false loading.value = false
return
} }
await callReplayStage()
} }
async function callGpuDirect() { /** Browser-side CV — no network, instant */
const body: Record<string, unknown> = { async function runLocal() {
image: props.frameImage, const t0 = performance.now()
// Decode base64 JPEG → ImageData
const imageData = await b64ToImageData(props.frameImage!)
// Build params from slider values
const params: Partial<EdgeDetectionParams> = {
cannyLow: values.value['edge_canny_low'] as number,
cannyHigh: values.value['edge_canny_high'] as number,
houghThreshold: values.value['edge_hough_threshold'] as number,
houghMinLength: values.value['edge_hough_min_length'] as number,
houghMaxGap: values.value['edge_hough_max_gap'] as number,
pairMaxDistance: values.value['edge_pair_max_distance'] as number,
pairMinDistance: values.value['edge_pair_min_distance'] as number,
} }
// Pass current slider values as edge detection params
const frameKey = String(props.frameRef ?? 0)
if (debugEnabled.value) {
const result = await runEdgeDetectionDebug(imageData, params)
execTimeMs.value = Math.round(performance.now() - t0)
regionCount.value = result.regions.length
// Convert ImageData overlays to base64 for FrameRenderer
const edgeB64 = await imageDataToB64(result.edgeImageData)
const linesB64 = await imageDataToB64(result.linesImageData)
emit('replay-result', {
regions_by_frame: { [frameKey]: result.regions },
debug: {
[frameKey]: {
edge_overlay_b64: edgeB64,
lines_overlay_b64: linesB64,
horizontal_count: result.horizontalCount,
pair_count: result.pairCount,
},
},
})
} else {
const result = await runEdgeDetection(imageData, params)
execTimeMs.value = Math.round(performance.now() - t0)
regionCount.value = result.regions.length
emit('replay-result', {
regions_by_frame: { [frameKey]: result.regions },
debug: {},
})
}
}
/** Server-side CV — calls GPU box via proxy */
async function runServer() {
const t0 = performance.now()
const body: Record<string, unknown> = { image: props.frameImage }
for (const f of fields.value) { for (const f of fields.value) {
if (f.name !== 'enabled') { if (f.name !== 'enabled') {
body[f.name] = values.value[f.name] body[f.name] = values.value[f.name]
@@ -109,7 +197,6 @@ async function callGpuDirect() {
? '/api/detect/gpu/detect_edges/debug' ? '/api/detect/gpu/detect_edges/debug'
: '/api/detect/gpu/detect_edges' : '/api/detect/gpu/detect_edges'
try {
const resp = await fetch(endpoint, { const resp = await fetch(endpoint, {
method: 'POST', method: 'POST',
headers: { 'Content-Type': 'application/json' }, headers: { 'Content-Type': 'application/json' },
@@ -118,16 +205,15 @@ async function callGpuDirect() {
if (!resp.ok) { if (!resp.ok) {
const detail = await resp.text() const detail = await resp.text()
error.value = `GPU call failed: ${detail}` throw new Error(detail)
return
} }
const data = await resp.json() const data = await resp.json()
execTimeMs.value = Math.round(performance.now() - t0)
regionCount.value = data.regions?.length ?? 0 regionCount.value = data.regions?.length ?? 0
// Build result in the same shape the parent expects
const frameKey = String(props.frameRef ?? 0) const frameKey = String(props.frameRef ?? 0)
const result: Record<string, unknown> = { const result: any = {
regions_by_frame: { [frameKey]: data.regions ?? [] }, regions_by_frame: { [frameKey]: data.regions ?? [] },
debug: {}, debug: {},
} }
@@ -138,69 +224,10 @@ async function callGpuDirect() {
lines_overlay_b64: data.lines_overlay_b64 ?? '', lines_overlay_b64: data.lines_overlay_b64 ?? '',
horizontal_count: data.horizontal_count ?? 0, horizontal_count: data.horizontal_count ?? 0,
pair_count: data.pair_count ?? 0, pair_count: data.pair_count ?? 0,
},
} }
} }
}
emit('replay-result', result as any)
} catch (e) {
error.value = `GPU call failed: ${e}`
} finally {
loading.value = false
}
}
async function callReplayStage() {
const overrides: Record<string, unknown> = {}
for (const f of fields.value) {
if (values.value[f.name] !== f.default) {
overrides[f.name] = values.value[f.name]
}
}
const overrideKey = stageToOverrideKey(props.stage)
const configOverrides = Object.keys(overrides).length > 0
? { [overrideKey]: overrides }
: null
const body = {
job_id: props.jobId,
stage: props.stage,
frame_refs: props.frameRef != null ? [props.frameRef] : null,
config_overrides: configOverrides,
debug: debugEnabled.value,
}
try {
const resp = await fetch('/api/detect/replay-stage', {
method: 'POST',
headers: { 'Content-Type': 'application/json' },
body: JSON.stringify(body),
})
if (!resp.ok) {
const detail = await resp.text()
error.value = `Replay failed: ${detail}`
return
}
const result = await resp.json()
regionCount.value = result.region_count ?? 0
emit('replay-result', result) emit('replay-result', result)
} catch (e) {
error.value = `Replay failed: ${e}`
} finally {
loading.value = false
}
}
function stageToOverrideKey(stage: string): string {
const map: Record<string, string> = {
detect_edges: 'region_analysis',
detect_objects: 'detection',
run_ocr: 'ocr',
match_brands: 'resolver',
}
return map[stage] || stage
} }
</script> </script>
@@ -211,6 +238,18 @@ function stageToOverrideKey(stage: string): string {
<button class="sliders-reset" @click="resetDefaults" title="Reset to defaults">Reset</button> <button class="sliders-reset" @click="resetDefaults" title="Reset to defaults">Reset</button>
</div> </div>
<!-- Local / Server toggle -->
<div class="mode-toggle">
<button
:class="['mode-btn', { active: execMode === 'local' }]"
@click="execMode = 'local'"
>Local</button>
<button
:class="['mode-btn', { active: execMode === 'server' }]"
@click="execMode = 'server'"
>Server</button>
</div>
<div v-if="error" class="sliders-error">{{ error }}</div> <div v-if="error" class="sliders-error">{{ error }}</div>
<div class="sliders-list"> <div class="sliders-list">
@@ -245,20 +284,25 @@ function stageToOverrideKey(stage: string): string {
</div> </div>
</div> </div>
<!-- Debug overlay toggle --> <!-- Footer -->
<label class="slider-field bool-field debug-toggle">
<input type="checkbox" v-model="debugEnabled" @change="onSliderChange" />
<span class="field-label">Show edge overlays</span>
</label>
<!-- Feedback -->
<div class="sliders-footer"> <div class="sliders-footer">
<button class="apply-btn" :disabled="loading" @click="applyReplay"> <button
class="apply-btn"
:disabled="loading || autoApply"
@click="applyDetection"
>
{{ loading ? 'Running...' : 'Apply' }} {{ loading ? 'Running...' : 'Apply' }}
</button> </button>
<label class="auto-apply-toggle">
<input type="checkbox" v-model="autoApply" />
<span>Auto</span>
</label>
<span v-if="regionCount != null" class="region-count"> <span v-if="regionCount != null" class="region-count">
{{ regionCount }} regions {{ regionCount }} regions
</span> </span>
<span v-if="execTimeMs != null" class="exec-time">
{{ execTimeMs }}ms
</span>
</div> </div>
</div> </div>
</template> </template>
@@ -298,6 +342,34 @@ function stageToOverrideKey(stage: string): string {
} }
.sliders-reset:hover { background: var(--surface-2); } .sliders-reset:hover { background: var(--surface-2); }
.mode-toggle {
display: flex;
gap: 1px;
background: var(--surface-2);
border-radius: 3px;
overflow: hidden;
}
.mode-btn {
flex: 1;
padding: 3px 8px;
border: none;
background: var(--surface-2);
color: var(--text-dim);
font-family: var(--font-mono);
font-size: 9px;
cursor: pointer;
transition: all 0.15s;
}
.mode-btn.active {
background: var(--surface-3);
color: var(--text-primary);
font-weight: 600;
}
.mode-btn:hover:not(.active) {
color: var(--text-secondary);
}
.sliders-error { .sliders-error {
color: var(--status-error); color: var(--status-error);
font-size: 10px; font-size: 10px;
@@ -354,7 +426,6 @@ function stageToOverrideKey(stage: string): string {
color: var(--text-dim); color: var(--text-dim);
} }
/* Range slider styling */
input[type="range"] { input[type="range"] {
-webkit-appearance: none; -webkit-appearance: none;
appearance: none; appearance: none;
@@ -388,11 +459,6 @@ input[type="checkbox"] {
accent-color: #00bcd4; accent-color: #00bcd4;
} }
.debug-toggle {
padding-top: var(--space-1);
border-top: var(--panel-border);
}
.sliders-footer { .sliders-footer {
display: flex; display: flex;
align-items: center; align-items: center;
@@ -415,8 +481,25 @@ input[type="checkbox"] {
.apply-btn:hover { opacity: 0.85; } .apply-btn:hover { opacity: 0.85; }
.apply-btn:disabled { opacity: 0.4; cursor: not-allowed; } .apply-btn:disabled { opacity: 0.4; cursor: not-allowed; }
.auto-apply-toggle {
display: flex;
align-items: center;
gap: 4px;
font-size: 9px;
color: var(--text-secondary);
cursor: pointer;
}
.auto-apply-toggle input { accent-color: #00bcd4; }
.region-count { .region-count {
color: var(--text-secondary); color: var(--text-secondary);
font-size: 10px; font-size: 10px;
} }
.exec-time {
color: var(--text-dim);
font-size: 9px;
margin-left: auto;
}
</style> </style>

View File

@@ -1,5 +1,5 @@
<script setup lang="ts"> <script setup lang="ts">
import { ref, computed } from 'vue' import { ref, computed, watch } from 'vue'
import { Panel } from 'mpr-ui-framework' import { Panel } from 'mpr-ui-framework'
import FrameRenderer from 'mpr-ui-framework/src/renderers/FrameRenderer.vue' import FrameRenderer from 'mpr-ui-framework/src/renderers/FrameRenderer.vue'
import type { FrameBBox, FrameOverlay } from 'mpr-ui-framework/src/renderers/FrameRenderer.vue' import type { FrameBBox, FrameOverlay } from 'mpr-ui-framework/src/renderers/FrameRenderer.vue'
@@ -10,9 +10,18 @@ const props = defineProps<{
status?: 'idle' | 'live' | 'processing' | 'error' status?: 'idle' | 'live' | 'processing' | 'error'
/** Debug overlay layers passed from parent (editor mode) */ /** Debug overlay layers passed from parent (editor mode) */
overlays?: FrameOverlay[] overlays?: FrameOverlay[]
/** Frame image from checkpoint (scenario mode) — overrides SSE */
frameImage?: string | null
/** Boxes from editor (local CV or server) — merged with SSE boxes */
editorBoxes?: import('mpr-ui-framework/src/renderers/FrameRenderer.vue').FrameBBox[]
}>() }>()
const imageSrc = ref('') const imageSrc = ref(props.frameImage ?? '')
// Sync prop → internal ref when checkpoint frame changes
watch(() => props.frameImage, (v) => {
if (v) imageSrc.value = v
})
// Per-stage box accumulation // Per-stage box accumulation
const stageBoxes = ref<Record<string, FrameBBox[]>>({}) const stageBoxes = ref<Record<string, FrameBBox[]>>({})
@@ -117,14 +126,19 @@ function sourceToStage(source: string): string {
return map[source] || 'match_brands' return map[source] || 'match_brands'
} }
// Filtered boxes — show all toggled-on stages // Filtered boxes — show all toggled-on stages + editor boxes
const visibleBoxes = computed<FrameBBox[]>(() => { const visibleBoxes = computed<FrameBBox[]>(() => {
const result: FrameBBox[] = [] const result: FrameBBox[] = []
// SSE boxes filtered by toggles
for (const [stage, boxes] of Object.entries(stageBoxes.value)) { for (const [stage, boxes] of Object.entries(stageBoxes.value)) {
if (activeToggles.value.has(stage)) { if (activeToggles.value.has(stage)) {
result.push(...boxes) result.push(...boxes)
} }
} }
// Editor boxes (from local CV or server) — always shown
if (props.editorBoxes?.length) {
result.push(...props.editorBoxes)
}
return result return result
}) })

View File

@@ -7,7 +7,7 @@ import type { DataSource } from 'mpr-ui-framework'
import { usePipelineStore } from '../stores/pipeline' import { usePipelineStore } from '../stores/pipeline'
const PIPELINE_NODES = [ const PIPELINE_NODES = [
'extract_frames', 'filter_scenes', 'detect_objects', 'preprocess', 'extract_frames', 'filter_scenes', 'detect_edges', 'detect_objects', 'preprocess',
'run_ocr', 'match_brands', 'escalate_vlm', 'escalate_cloud', 'compile_report', 'run_ocr', 'match_brands', 'escalate_vlm', 'escalate_cloud', 'compile_report',
] ]

View File

@@ -0,0 +1,278 @@
/**
* Edge detection — TypeScript port of gpu/models/cv/edges.py
*
* 1:1 with the Python version. Same algorithm, same parameters,
* same output format. Runs in the browser, no network.
*/
import { toGrayscale, canny } from './imageOps'
import { houghLinesP, type LineSegment } from './hough'
export interface EdgeRegion {
x: number
y: number
w: number
h: number
confidence: number
label: string
}
export interface EdgeDetectionParams {
cannyLow: number
cannyHigh: number
houghThreshold: number
houghMinLength: number
houghMaxGap: number
pairMaxDistance: number
pairMinDistance: number
}
export interface EdgeDetectionResult {
regions: EdgeRegion[]
}
export interface EdgeDetectionDebugResult extends EdgeDetectionResult {
edgeImageData: ImageData // Canny output for overlay
linesImageData: ImageData // Frame with Hough lines drawn
horizontalCount: number
pairCount: number
}
type HLine = { xMin: number; xMax: number; yMid: number; length: number }
/** Set a pixel on ImageData with bounds check */
function setPixel(img: ImageData, x: number, y: number, r: number, g: number, b: number) {
if (x >= 0 && x < img.width && y >= 0 && y < img.height) {
const p = (y * img.width + x) * 4
img.data[p] = r
img.data[p + 1] = g
img.data[p + 2] = b
img.data[p + 3] = 255
}
}
/** Bresenham line drawing with thickness */
function drawLineThick(
img: ImageData,
x0: number, y0: number, x1: number, y1: number,
r: number, g: number, b: number,
thickness: number = 1,
) {
const dx = Math.abs(x1 - x0)
const dy = Math.abs(y1 - y0)
const sx = x0 < x1 ? 1 : -1
const sy = y0 < y1 ? 1 : -1
let err = dx - dy
const half = Math.floor(thickness / 2)
while (true) {
for (let oy = -half; oy <= half; oy++) {
for (let ox = -half; ox <= half; ox++) {
setPixel(img, x0 + ox, y0 + oy, r, g, b)
}
}
if (x0 === x1 && y0 === y1) break
const e2 = 2 * err
if (e2 > -dy) { err -= dy; x0 += sx }
if (e2 < dx) { err += dx; y0 += sy }
}
}
const DEFAULT_PARAMS: EdgeDetectionParams = {
cannyLow: 50,
cannyHigh: 150,
houghThreshold: 80,
houghMinLength: 100,
houghMaxGap: 10,
pairMaxDistance: 200,
pairMinDistance: 15,
}
/** Filter to near-horizontal lines (within 10 degrees) */
function filterHorizontal(lines: LineSegment[], maxAngleDeg: number = 10): HLine[] {
const maxSlope = Math.tan((maxAngleDeg * Math.PI) / 180)
const result: HLine[] = []
for (const line of lines) {
const dx = line.x2 - line.x1
if (dx === 0) continue
const slope = Math.abs((line.y2 - line.y1) / dx)
if (slope <= maxSlope) {
const yMid = (line.y1 + line.y2) / 2
const xMin = Math.min(line.x1, line.x2)
const xMax = Math.max(line.x1, line.x2)
const length = Math.sqrt(dx * dx + (line.y2 - line.y1) ** 2)
result.push({ xMin, xMax, yMid, length })
}
}
return result
}
/** Find pairs of horizontal lines that could be top/bottom of a hoarding */
function findLinePairs(
horizontals: HLine[],
minDistance: number,
maxDistance: number,
): [HLine, HLine][] {
const sorted = [...horizontals].sort((a, b) => a.yMid - b.yMid)
const pairs: [HLine, HLine][] = []
const used = new Set<number>()
for (let i = 0; i < sorted.length; i++) {
if (used.has(i)) continue
const top = sorted[i]
for (let j = i + 1; j < sorted.length; j++) {
if (used.has(j)) continue
const bottom = sorted[j]
const yGap = bottom.yMid - top.yMid
if (yGap < minDistance) continue
if (yGap > maxDistance) break
// Check horizontal overlap (50% of shorter line)
const overlapStart = Math.max(top.xMin, bottom.xMin)
const overlapEnd = Math.min(top.xMax, bottom.xMax)
const overlap = overlapEnd - overlapStart
const shorterLength = Math.min(top.xMax - top.xMin, bottom.xMax - bottom.xMin)
if (shorterLength > 0 && overlap / shorterLength >= 0.5) {
pairs.push([top, bottom])
used.add(i)
used.add(j)
break
}
}
}
return pairs
}
/** Convert a line pair to a bounding box */
function pairToBox(
top: HLine,
bottom: HLine,
frameWidth: number,
frameHeight: number,
): EdgeRegion | null {
const x = Math.max(0, Math.min(top.xMin, bottom.xMin))
const y = Math.max(0, top.yMid)
const x2 = Math.min(frameWidth, Math.max(top.xMax, bottom.xMax))
const y2 = Math.min(frameHeight, bottom.yMid)
const w = x2 - x
const h = y2 - y
if (w < 20 || h < 5) return null
const avgLineLength = (top.length + bottom.length) / 2
const coverage = Math.min(1.0, avgLineLength / Math.max(w, 1))
return {
x: Math.round(x),
y: Math.round(y),
w: Math.round(w),
h: Math.round(h),
confidence: Math.round(coverage * 1000) / 1000,
label: 'edge_region',
}
}
/**
* Detect edges in an RGBA ImageData.
*
* Equivalent to gpu/models/cv/edges.py detect_edges()
*/
export function detectEdges(
imageData: ImageData,
params: Partial<EdgeDetectionParams> = {},
): EdgeDetectionResult {
const p = { ...DEFAULT_PARAMS, ...params }
const { width, height } = imageData
const gray = toGrayscale(imageData.data, width, height)
const edges = canny(gray, width, height, p.cannyLow, p.cannyHigh)
const rawLines = houghLinesP(edges, width, height, p.houghThreshold, p.houghMinLength, p.houghMaxGap)
const horizontals = filterHorizontal(rawLines)
if (horizontals.length < 2) return { regions: [] }
const pairs = findLinePairs(horizontals, p.pairMinDistance, p.pairMaxDistance)
const regions: EdgeRegion[] = []
for (const [top, bottom] of pairs) {
const box = pairToBox(top, bottom, width, height)
if (box) regions.push(box)
}
return { regions }
}
/**
* Detect edges with debug visualizations.
*
* Equivalent to gpu/models/cv/edges.py detect_edges_debug()
*/
export function detectEdgesDebug(
imageData: ImageData,
params: Partial<EdgeDetectionParams> = {},
): EdgeDetectionDebugResult {
const p = { ...DEFAULT_PARAMS, ...params }
const { width, height, data } = imageData
const gray = toGrayscale(data, width, height)
const edges = canny(gray, width, height, p.cannyLow, p.cannyHigh)
// Edge overlay — white edges on black
const edgeImageData = new ImageData(width, height)
for (let i = 0; i < edges.length; i++) {
const px = i * 4
edgeImageData.data[px] = edges[i]
edgeImageData.data[px + 1] = edges[i]
edgeImageData.data[px + 2] = edges[i]
edgeImageData.data[px + 3] = 255
}
const rawLines = houghLinesP(edges, width, height, p.houghThreshold, p.houghMinLength, p.houghMaxGap)
const horizontals = filterHorizontal(rawLines)
// Lines overlay — darken original frame so lines pop, then draw
const linesImageData = new ImageData(new Uint8ClampedArray(data), width, height)
for (let i = 0; i < linesImageData.data.length; i += 4) {
linesImageData.data[i] = Math.round(linesImageData.data[i] * 0.3)
linesImageData.data[i + 1] = Math.round(linesImageData.data[i + 1] * 0.3)
linesImageData.data[i + 2] = Math.round(linesImageData.data[i + 2] * 0.3)
}
// Draw all Hough lines in red (3px thick)
for (const line of rawLines) {
drawLineThick(linesImageData, line.x1, line.y1, line.x2, line.y2, 255, 50, 50, 2)
}
// Draw horizontal lines in cyan (3px thick)
for (const h of horizontals) {
drawLineThick(linesImageData, Math.round(h.xMin), Math.round(h.yMid), Math.round(h.xMax), Math.round(h.yMid), 0, 255, 255, 3)
}
const pairs = horizontals.length >= 2
? findLinePairs(horizontals, p.pairMinDistance, p.pairMaxDistance)
: []
// Draw paired lines in bright green (4px thick)
for (const [top, bottom] of pairs) {
drawLineThick(linesImageData, Math.round(top.xMin), Math.round(top.yMid), Math.round(top.xMax), Math.round(top.yMid), 0, 255, 0, 4)
drawLineThick(linesImageData, Math.round(bottom.xMin), Math.round(bottom.yMid), Math.round(bottom.xMax), Math.round(bottom.yMid), 0, 255, 0, 4)
}
const regions: EdgeRegion[] = []
for (const [top, bottom] of pairs) {
const box = pairToBox(top, bottom, width, height)
if (box) regions.push(box)
}
return {
regions,
edgeImageData,
linesImageData,
horizontalCount: horizontals.length,
pairCount: pairs.length,
}
}

View File

@@ -0,0 +1,147 @@
/**
* Probabilistic Hough Line Transform — pure TypeScript.
*
* Equivalent to cv2.HoughLinesP. Finds line segments in a binary edge image.
*/
export interface LineSegment {
x1: number
y1: number
x2: number
y2: number
}
/**
* Probabilistic Hough Line Transform.
*
* @param edges - Binary edge image (255 = edge, 0 = not)
* @param width - Image width
* @param height - Image height
* @param threshold - Accumulator threshold (min votes for a line)
* @param minLineLength - Minimum line length in pixels
* @param maxLineGap - Maximum gap between points on the same line
*/
export function houghLinesP(
edges: Uint8Array,
width: number,
height: number,
threshold: number,
minLineLength: number,
maxLineGap: number,
): LineSegment[] {
const diag = Math.ceil(Math.sqrt(width * width + height * height))
const numAngles = 180
const rhoMax = diag
// Precompute sin/cos tables
const cosTable = new Float64Array(numAngles)
const sinTable = new Float64Array(numAngles)
for (let t = 0; t < numAngles; t++) {
const angle = (t * Math.PI) / numAngles
cosTable[t] = Math.cos(angle)
sinTable[t] = Math.sin(angle)
}
// Collect edge points
const edgePoints: [number, number][] = []
for (let y = 0; y < height; y++) {
for (let x = 0; x < width; x++) {
if (edges[y * width + x] === 255) {
edgePoints.push([x, y])
}
}
}
if (edgePoints.length === 0) return []
// Shuffle edge points for probabilistic sampling
for (let i = edgePoints.length - 1; i > 0; i--) {
const j = Math.floor(Math.random() * (i + 1))
const tmp = edgePoints[i]
edgePoints[i] = edgePoints[j]
edgePoints[j] = tmp
}
// Accumulator
const accum = new Int32Array(numAngles * (2 * rhoMax + 1))
const used = new Uint8Array(width * height)
const lines: LineSegment[] = []
for (const [px, py] of edgePoints) {
if (used[py * width + px]) continue
// Vote
let maxVotes = 0
let bestTheta = 0
for (let t = 0; t < numAngles; t++) {
const rho = Math.round(px * cosTable[t] + py * sinTable[t]) + rhoMax
const idx = t * (2 * rhoMax + 1) + rho
accum[idx]++
if (accum[idx] > maxVotes) {
maxVotes = accum[idx]
bestTheta = t
}
}
if (maxVotes < threshold) continue
// Walk along the line at bestTheta through (px, py)
const ct = cosTable[bestTheta]
const st = sinTable[bestTheta]
// Line direction is perpendicular to (cos, sin)
const dx = -st
const dy = ct
// Walk forward and backward to find line extent
const walkLine = (startX: number, startY: number, dirX: number, dirY: number): [number, number] => {
let lastEdgeX = startX
let lastEdgeY = startY
let gap = 0
let cx = startX
let cy = startY
for (let step = 1; step < Math.max(width, height); step++) {
const nx = Math.round(cx + dirX * step)
const ny = Math.round(cy + dirY * step)
if (nx < 0 || nx >= width || ny < 0 || ny >= height) break
if (edges[ny * width + nx] === 255 && !used[ny * width + nx]) {
lastEdgeX = nx
lastEdgeY = ny
gap = 0
} else {
gap++
if (gap > maxLineGap) break
}
}
return [lastEdgeX, lastEdgeY]
}
const [x1, y1] = walkLine(px, py, -dx, -dy)
const [x2, y2] = walkLine(px, py, dx, dy)
const length = Math.sqrt((x2 - x1) ** 2 + (y2 - y1) ** 2)
if (length < minLineLength) continue
// Mark pixels as used
const steps = Math.ceil(length)
for (let s = 0; s <= steps; s++) {
const mx = Math.round(x1 + (x2 - x1) * s / steps)
const my = Math.round(y1 + (y2 - y1) * s / steps)
if (mx >= 0 && mx < width && my >= 0 && my < height) {
used[my * width + mx] = 1
}
}
lines.push({ x1, y1, x2, y2 })
// Unvote (clean accumulator for used points)
for (let t = 0; t < numAngles; t++) {
const rho = Math.round(px * cosTable[t] + py * sinTable[t]) + rhoMax
accum[t * (2 * rhoMax + 1) + rho]--
}
}
return lines
}

View File

@@ -0,0 +1,190 @@
/**
* Pure TypeScript image operations — no OpenCV dependency.
*
* These implement the subset of CV operations needed for the edge
* detection editor. Same algorithms as the Python/OpenCV versions
* but running in the browser with zero WASM overhead.
*/
/** Grayscale from RGBA ImageData */
export function toGrayscale(data: Uint8ClampedArray, width: number, height: number): Uint8Array {
const gray = new Uint8Array(width * height)
for (let i = 0; i < gray.length; i++) {
const p = i * 4
// ITU-R BT.601 luma
gray[i] = Math.round(0.299 * data[p] + 0.587 * data[p + 1] + 0.114 * data[p + 2])
}
return gray
}
/** 5x5 Gaussian blur */
export function gaussianBlur(src: Uint8Array, width: number, height: number): Uint8Array {
// 5x5 Gaussian kernel (sigma ~1.4, matches OpenCV default for Canny)
const kernel = [
2, 4, 5, 4, 2,
4, 9, 12, 9, 4,
5, 12, 15, 12, 5,
4, 9, 12, 9, 4,
2, 4, 5, 4, 2,
]
const kSum = 159
const out = new Uint8Array(width * height)
for (let y = 2; y < height - 2; y++) {
for (let x = 2; x < width - 2; x++) {
let sum = 0
for (let ky = -2; ky <= 2; ky++) {
for (let kx = -2; kx <= 2; kx++) {
sum += src[(y + ky) * width + (x + kx)] * kernel[(ky + 2) * 5 + (kx + 2)]
}
}
out[y * width + x] = Math.round(sum / kSum)
}
}
return out
}
/** Sobel gradients → magnitude + direction */
export function sobelGradients(
src: Uint8Array,
width: number,
height: number,
): { magnitude: Float32Array; direction: Float32Array } {
const size = width * height
const magnitude = new Float32Array(size)
const direction = new Float32Array(size)
for (let y = 1; y < height - 1; y++) {
for (let x = 1; x < width - 1; x++) {
const i = y * width + x
// Sobel 3x3
const gx =
-src[(y - 1) * width + (x - 1)] - 2 * src[y * width + (x - 1)] - src[(y + 1) * width + (x - 1)] +
src[(y - 1) * width + (x + 1)] + 2 * src[y * width + (x + 1)] + src[(y + 1) * width + (x + 1)]
const gy =
-src[(y - 1) * width + (x - 1)] - 2 * src[(y - 1) * width + x] - src[(y - 1) * width + (x + 1)] +
src[(y + 1) * width + (x - 1)] + 2 * src[(y + 1) * width + x] + src[(y + 1) * width + (x + 1)]
magnitude[i] = Math.sqrt(gx * gx + gy * gy)
direction[i] = Math.atan2(gy, gx)
}
}
return { magnitude, direction }
}
/** Non-maximum suppression for Canny */
export function nonMaxSuppression(
magnitude: Float32Array,
direction: Float32Array,
width: number,
height: number,
): Float32Array {
const out = new Float32Array(width * height)
for (let y = 1; y < height - 1; y++) {
for (let x = 1; x < width - 1; x++) {
const i = y * width + x
const mag = magnitude[i]
if (mag === 0) continue
// Quantize direction to 4 angles (0, 45, 90, 135)
let angle = (direction[i] * 180) / Math.PI
if (angle < 0) angle += 180
let n1 = 0, n2 = 0
if ((angle < 22.5) || (angle >= 157.5)) {
n1 = magnitude[y * width + (x + 1)]
n2 = magnitude[y * width + (x - 1)]
} else if (angle < 67.5) {
n1 = magnitude[(y - 1) * width + (x + 1)]
n2 = magnitude[(y + 1) * width + (x - 1)]
} else if (angle < 112.5) {
n1 = magnitude[(y - 1) * width + x]
n2 = magnitude[(y + 1) * width + x]
} else {
n1 = magnitude[(y - 1) * width + (x - 1)]
n2 = magnitude[(y + 1) * width + (x + 1)]
}
out[i] = (mag >= n1 && mag >= n2) ? mag : 0
}
}
return out
}
/** Hysteresis thresholding for Canny */
export function hysteresis(
nms: Float32Array,
width: number,
height: number,
low: number,
high: number,
): Uint8Array {
const out = new Uint8Array(width * height)
// Mark strong and weak edges
const STRONG = 255
const WEAK = 128
for (let i = 0; i < nms.length; i++) {
if (nms[i] >= high) out[i] = STRONG
else if (nms[i] >= low) out[i] = WEAK
}
// Connect weak edges adjacent to strong edges
let changed = true
while (changed) {
changed = false
for (let y = 1; y < height - 1; y++) {
for (let x = 1; x < width - 1; x++) {
const i = y * width + x
if (out[i] !== WEAK) continue
// Check 8-neighbors for strong edge
for (let dy = -1; dy <= 1; dy++) {
for (let dx = -1; dx <= 1; dx++) {
if (out[(y + dy) * width + (x + dx)] === STRONG) {
out[i] = STRONG
changed = true
}
}
}
}
}
}
// Suppress remaining weak edges
for (let i = 0; i < out.length; i++) {
if (out[i] !== STRONG) out[i] = 0
}
return out
}
/** Full Canny edge detection */
export function canny(
gray: Uint8Array,
width: number,
height: number,
lowThreshold: number,
highThreshold: number,
): Uint8Array {
const blurred = gaussianBlur(gray, width, height)
const { magnitude, direction } = sobelGradients(blurred, width, height)
const nms = nonMaxSuppression(magnitude, direction, width, height)
return hysteresis(nms, width, height, lowThreshold, highThreshold)
}
/** Convert edge image (Uint8Array) to base64 JPEG via offscreen canvas */
export function edgeImageToB64(edges: Uint8Array, width: number, height: number): string {
const canvas = new OffscreenCanvas(width, height)
const ctx = canvas.getContext('2d')!
const imgData = ctx.createImageData(width, height)
for (let i = 0; i < edges.length; i++) {
const p = i * 4
imgData.data[p] = edges[i]
imgData.data[p + 1] = edges[i]
imgData.data[p + 2] = edges[i]
imgData.data[p + 3] = 255
}
ctx.putImageData(imgData, 0, 0)
const blob = canvas.convertToBlob({ type: 'image/jpeg', quality: 0.7 })
return '' // placeholder — async handled in worker
}

View File

@@ -0,0 +1,80 @@
/**
* Browser-side CV — public API.
*
* Runs edge detection directly on the main thread.
* Pure TypeScript, no WASM, no dependencies.
* ~10-50ms per 1080p frame — fast enough for slider feedback.
*
* TODO: Move to Web Worker when processing larger batches.
*
* Usage:
* import { runEdgeDetection, runEdgeDetectionDebug } from 'mpr-ui-framework/src/cv'
* const result = await runEdgeDetection(imageData, params)
*/
import { detectEdges, detectEdgesDebug, type EdgeRegion, type EdgeDetectionParams } from './edges'
export type { EdgeRegion, EdgeDetectionParams } from './edges'
export type { EdgeDetectionResult, EdgeDetectionDebugResult } from './edges'
/** Run edge detection. Returns bounding boxes. */
export async function runEdgeDetection(
imageData: ImageData,
params: Partial<EdgeDetectionParams> = {},
): Promise<{ regions: EdgeRegion[] }> {
return detectEdges(imageData, params)
}
/** Run edge detection with debug overlays. Returns boxes + visualization ImageData. */
export async function runEdgeDetectionDebug(
imageData: ImageData,
params: Partial<EdgeDetectionParams> = {},
): Promise<{
regions: EdgeRegion[]
edgeImageData: ImageData
linesImageData: ImageData
horizontalCount: number
pairCount: number
}> {
return detectEdgesDebug(imageData, params)
}
/**
* Decode a base64 JPEG string to ImageData.
*
* Used to convert the checkpoint frame (base64) into ImageData
* that the CV functions can process.
*/
export function b64ToImageData(b64: string): Promise<ImageData> {
return new Promise((resolve, reject) => {
const img = new Image()
img.onload = () => {
const canvas = new OffscreenCanvas(img.width, img.height)
const ctx = canvas.getContext('2d')!
ctx.drawImage(img, 0, 0)
resolve(ctx.getImageData(0, 0, img.width, img.height))
}
img.onerror = () => reject(new Error('Failed to decode image'))
img.src = `data:image/jpeg;base64,${b64}`
})
}
/**
* Encode ImageData to base64 JPEG string.
*
* Used to convert debug overlay ImageData back to base64
* for the FrameRenderer overlays prop.
*/
export async function imageDataToB64(imageData: ImageData): Promise<string> {
const canvas = new OffscreenCanvas(imageData.width, imageData.height)
const ctx = canvas.getContext('2d')!
ctx.putImageData(imageData, 0, 0)
const blob = await canvas.convertToBlob({ type: 'image/jpeg', quality: 0.7 })
const buffer = await blob.arrayBuffer()
const bytes = new Uint8Array(buffer)
let binary = ''
for (let i = 0; i < bytes.length; i++) {
binary += String.fromCharCode(bytes[i])
}
return btoa(binary)
}

View File

@@ -0,0 +1,41 @@
/**
* CV Web Worker — runs edge detection off the main thread.
*
* Message protocol:
* Main → Worker: { type: 'detect_edges', imageData: ImageData, params: {...} }
* Main → Worker: { type: 'detect_edges_debug', imageData: ImageData, params: {...} }
* Worker → Main: { type: 'result', regions: [...] }
* Worker → Main: { type: 'debug_result', regions: [...], edgeImageData, linesImageData, horizontalCount, pairCount }
* Worker → Main: { type: 'error', message: string }
*/
import { detectEdges, detectEdgesDebug, type EdgeDetectionParams } from './edges'
self.onmessage = (event: MessageEvent) => {
const { type, imageData, params } = event.data
try {
if (type === 'detect_edges') {
const result = detectEdges(imageData, params)
self.postMessage({ type: 'result', regions: result.regions })
} else if (type === 'detect_edges_debug') {
const result = detectEdgesDebug(imageData, params)
self.postMessage({
type: 'debug_result',
regions: result.regions,
edgeImageData: result.edgeImageData,
linesImageData: result.linesImageData,
horizontalCount: result.horizontalCount,
pairCount: result.pairCount,
}, [
// Transfer ownership of the backing buffers for zero-copy
result.edgeImageData.data.buffer,
result.linesImageData.data.buffer,
])
} else {
self.postMessage({ type: 'error', message: `Unknown message type: ${type}` })
}
} catch (e) {
self.postMessage({ type: 'error', message: String(e) })
}
}

View File

@@ -41,7 +41,7 @@ const flowNodes = computed(() =>
label: n.id.replace(/_/g, ' '), label: n.id.replace(/_/g, ' '),
status: n.status, status: n.status,
color: statusColors[n.status] ?? statusColors.pending, color: statusColors[n.status] ?? statusColors.pending,
textColor: n.status === 'pending' ? '#888' : '#000', textColor: '#fff',
hasRegionEditor: regionStageSet.value.has(n.id), hasRegionEditor: regionStageSet.value.has(n.id),
isRunning: n.status === 'running', isRunning: n.status === 'running',
}, },