TZGyn/kapture
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 52 Wiki Activity

Retry OCR on upscaled images when nothing is read

Browse Source 
Tiny UI text (around 7pt) reads as nothing at native size — verified
locally. When a pass returns fewer than three characters, OCR retries at
2x and 3x and keeps the fullest result. The smoke test covers a tiny-text
case so the fallback stays working.

💘 Generated with Crush

Assisted-by: Crush:deepseek-v4-flash-0731
This commit is contained in:
TZGyn
2026-09-30 11:44:26 +08:00
parent 2e971d5f8d
commit 04b84e188a
3 changed files with 73 additions and 36 deletions
Download Patch File Download Diff File Expand all files Collapse all files
Sources/Kapture/OCRService.swift
+58 -26
View File
@@ -8,34 +8,66 @@ final class OCRService {
func recognize(_ cgImage: CGImage, completion: @escaping (String) -> Void) {
queue.async {
let request = VNRecognizeTextRequest()
request.recognitionLevel = .accurate
request.usesLanguageCorrection = true
request.recognitionLanguages = ["en-US", "zh-Hans", "zh-Hant", "ja-JP", "ko-KR"]
// Without this, Vision defaults to the first listed language and
// renders CJK as Latin garbage (verified: 你好世界 -> "177WR").
if #available(macOS 13.0, *) {
request.automaticallyDetectsLanguage = true
}
let handler = VNImageRequestHandler(cgImage: cgImage, options: [:])
do {
try handler.perform([request])
} catch {
completion("")
let text = Self.perform(cgImage)
// Tiny text (small UI labels) often reads as nothing at native
// size; retry upscaled and keep whichever pass read more.
guard text.count < 3 else {
completion(text)
return
}
let observations = request.results ?? []
let lines = observations
.sorted { a, b in
let ay = a.boundingBox.midY
let by = b.boundingBox.midY
if abs(ay - by) > 0.02 { return ay > by }
return a.boundingBox.minX < b.boundingBox.minX
}
.compactMap { $0.topCandidates(1).first?.string }
completion(lines.joined(separator: "\n"))
var best = text
for factor in [CGFloat(2), CGFloat(3)] {
guard let upscaled = Self.upscaled(cgImage, factor: factor) else { continue }
let candidate = Self.perform(upscaled)
if candidate.count > best.count { best = candidate }
if best.count >= 3 { break }
}
completion(best)
}
}
private static func perform(_ cgImage: CGImage) -> String {
let request = VNRecognizeTextRequest()
request.recognitionLevel = .accurate
request.usesLanguageCorrection = true
request.recognitionLanguages = ["en-US", "zh-Hans", "zh-Hant", "ja-JP", "ko-KR"]
// Without this, Vision defaults to the first listed language and
// renders CJK as Latin garbage (verified: 你好世界 -> "177WR").
if #available(macOS 13.0, *) {
request.automaticallyDetectsLanguage = true
}
let handler = VNImageRequestHandler(cgImage: cgImage, options: [:])
do {
try handler.perform([request])
} catch {
return ""
}
let observations = request.results ?? []
let lines = observations
.sorted { a, b in
let ay = a.boundingBox.midY
let by = b.boundingBox.midY
if abs(ay - by) > 0.02 { return ay > by }
return a.boundingBox.minX < b.boundingBox.minX
}
.compactMap { $0.topCandidates(1).first?.string }
return lines.joined(separator: "\n")
}
private static func upscaled(_ cgImage: CGImage, factor: CGFloat) -> CGImage? {
let width = Int(CGFloat(cgImage.width) * factor)
let height = Int(CGFloat(cgImage.height) * factor)
guard width > 0, height > 0, width <= 20000, height <= 20000 else { return nil }
guard let ctx = CGContext(
data: nil, width: width, height: height,
bitsPerComponent: 8, bytesPerRow: 0,
space: CGColorSpaceCreateDeviceRGB(),
bitmapInfo: CGImageAlphaInfo.premultipliedLast.rawValue
) else { return nil }
ctx.interpolationQuality = .high
ctx.draw(cgImage, in: CGRect(x: 0, y: 0, width: CGFloat(width), height: CGFloat(height)))
return ctx.makeImage()
}
}