当我在我的 Android 项目中实现 Google ML OCR Tesseract 时，没有获得正确的文本

Question

I am trying to implement Google ML OCR Tesseract in my android project and i am getting some random text from the picture I have clicked. Following is my code:

类 MainActivity : ComponentActivity() { 私有 Lateinit var tessBaseAPI：TessBaseAPI private Lateinit var imageView: ImageView private Lateinit var textView: TextView 私有val requestImageCapture = 1 私人 val takePicture = registerForActivityResult(ActivityResultContracts.StartActivityForResult()) { 结果 ->

        if (result.resultCode == RESULT_OK) {
            val resultInfo = result.data?.extras?.get("data")
            val imageBitmap = result.data?.extras?.get("data") as Bitmap?
            imageBitmap?.let {
                imageView.setImageBitmap(it)
                recognizeText(it)
            }
        }
    }

override fun onCreate(savedInstanceState: Bundle?) {
    super.onCreate(savedInstanceState)
    setContentView(R.layout.activity_main)

    imageView = findViewById(R.id.imageView)
    textView = findViewById(R.id.textView)
    val captureButton: Button = findViewById(R.id.captureButton)

    copyLanguageDataToInternalStorage()
    initTesseract()

    captureButton.setOnClickListener {
        dispatchTakePictureIntent()
    }
}

private fun initTesseract() {
    val datapath = "${filesDir}/tesseract/"
    val dir = File(datapath)
    if (!dir.exists()) {
        dir.mkdirs()
    }
    val datafilepath = "${datapath}/tessdata/"
    val filesDir = File(datafilepath)
    if (!filesDir.exists()) {
        filesDir.mkdirs()
    }
    Log.d("dataPath", "data: $datapath")
    tessBaseAPI = TessBaseAPI()
    tessBaseAPI.init(datapath, "eng")
}

private fun copyLanguageDataToInternalStorage() {
    Log.d("test", "testing")
    try {
        val assetManager = assets
        val inputStream = assetManager.open("tesseract/tessdata/eng.traineddata")
        val outputStream =
            FileOutputStream(File(filesDir, "tesseract/tessdata/eng.traineddata"))

        inputStream.use { input ->
            outputStream.use { output ->
                input.copyTo(output)
            }
        }

    } catch (e: IOException) {

        e.printStackTrace()
    }
}

private fun dispatchTakePictureIntent() {
    val takePictureIntent = Intent(MediaStore.ACTION_IMAGE_CAPTURE)
    if (takePictureIntent.resolveActivity(packageManager) != null) {
        takePicture.launch(takePictureIntent)
    }
}

private fun recognizeText(bitmap: Bitmap) {
    tessBaseAPI.setImage(bitmap)
    val recognizedText = tessBaseAPI.getUTF8Text()
    textView.text = recognizedText
}

override fun onDestroy() {
    super.onDestroy()
    tessBaseAPI.end()
}

}

and i am getting this as text

`m”;-.v__m-~_;;;;;;§'F"WWW If?`

Can Someone tell me where i am doing wrong in getting text from capturing a picture from the phone

Answer 1

您没有描述如何将 Tesseract 包含到您的项目中（通过包装器还是直接？），但 tesseract

setImage

需要 Leptonica PIX 对象进行识别，而不是位图。请检查这是否不是您问题的原因（例如通过 tesseract 变量

tessedit_write_images

）。

我不是 Android 开发人员，但是

initTesseract

函数的逻辑很可疑 - 它只是创建一个目录结构而不调用

copyLanguageDataToInternalStorage

- 这可能会导致你尝试在没有任何语言文件的情况下初始化 tesseract API，因此 API 会不被启动...