All files / web/src/app/api/vision-training/preview-masked route.ts

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
import { NextResponse } from 'next/server'
import { spawn } from 'child_process'
import path from 'path'
import { TRAINING_PYTHON, PYTHON_ENV, ensureVenvReady } from '../config'
import { withAuth } from '@/lib/auth/withAuth'

/**
 * Preview marker masking using the EXACT SAME Python code as training.
 *
 * This ensures the preview matches what the model will see during training.
 */
export const POST = withAuth(
  async (request) => {
    try {
      const body = await request.json()
      const { imageData, corners, method = 'noise' } = body

      if (!imageData || !corners) {
        return NextResponse.json({ error: 'Missing imageData or corners' }, { status: 400 })
      }

      // Validate corners format
      if (!corners.topLeft || !corners.topRight || !corners.bottomLeft || !corners.bottomRight) {
        return NextResponse.json(
          {
            error: 'Invalid corners format - need topLeft, topRight, bottomLeft, bottomRight',
          },
          { status: 400 }
        )
      }

      // Ensure venv is ready
      const venvResult = await ensureVenvReady()
      if (!venvResult.success) {
        return NextResponse.json(
          { error: `Python environment not ready: ${venvResult.error}` },
          { status: 500 }
        )
      }

      // Prepare input for Python script
      const pythonInput = JSON.stringify({
        image_base64: imageData,
        corners: [corners.topLeft, corners.topRight, corners.bottomLeft, corners.bottomRight],
        method,
      })

      // Path to the marker masking script
      const scriptPath = path.join(
        process.cwd(),
        'scripts',
        'train-boundary-detector',
        'marker_masking.py'
      )

      // Run Python script using the training venv
      // Pass input via stdin instead of command line to avoid E2BIG for large images
      const result = await new Promise<{
        masked_image_base64: string
        mask_regions: Array<{ x1: number; y1: number; x2: number; y2: number }>
      }>((resolve, reject) => {
        const pythonProcess = spawn(TRAINING_PYTHON, [scriptPath], {
          cwd: path.join(process.cwd(), 'scripts', 'train-boundary-detector'),
          env: PYTHON_ENV,
        })

        // Send input via stdin
        pythonProcess.stdin.write(pythonInput)
        pythonProcess.stdin.end()

        let stdout = ''
        let stderr = ''

        pythonProcess.stdout.on('data', (data) => {
          stdout += data.toString()
        })

        pythonProcess.stderr.on('data', (data) => {
          stderr += data.toString()
        })

        pythonProcess.on('close', (code) => {
          if (code !== 0) {
            reject(new Error(`Python script failed: ${stderr}`))
            return
          }

          try {
            const output = JSON.parse(stdout)
            resolve(output)
          } catch {
            reject(new Error(`Failed to parse Python output: ${stdout}`))
          }
        })

        pythonProcess.on('error', (err) => {
          reject(err)
        })
      })

      return NextResponse.json({
        success: true,
        maskedImageData: result.masked_image_base64,
        maskRegions: result.mask_regions,
      })
    } catch (error) {
      console.error('[preview-masked] Error:', error)
      return NextResponse.json(
        { error: error instanceof Error ? error.message : 'Unknown error' },
        { status: 500 }
      )
    }
  },
  { role: 'admin' }
)