Java TessBaseAPI类代码示例

import com.googlecode.tesseract.android.TessBaseAPI; //导入依赖的package包/类
public String detectText(Bitmap bitmap) {
    Log.d(TAG, "Initialization of TessBaseApi");
    TessDataManager.initTessTrainedData(context);
    TessBaseAPI tessBaseAPI = new TessBaseAPI();
    String path = TessDataManager.getTesseractFolder();
    Log.d(TAG, "Tess folder: " + path);
    tessBaseAPI.setDebug(true);
    tessBaseAPI.init(path, "eng");
    // 白名单
    tessBaseAPI.setVariable(TessBaseAPI.VAR_CHAR_WHITELIST, "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789");
    // 黑名单
    tessBaseAPI.setVariable(TessBaseAPI.VAR_CHAR_BLACKLIST, "[email protected]#$%^&*()_+=-[]}{;:'\"\\|~`,./<>?"); 
    tessBaseAPI.setPageSegMode(TessBaseAPI.PageSegMode.PSM_AUTO_OSD);
    Log.d(TAG, "Ended initialization of TessEngine");
    Log.d(TAG, "Running inspection on bitmap");
    tessBaseAPI.setImage(bitmap);
    String inspection = tessBaseAPI.getHOCRText(0);

    Log.d(TAG, "Confidence values: " + tessBaseAPI.meanConfidence());
    tessBaseAPI.end();
    System.gc();
    return Tools.getTelNum(inspection);
}
 

import com.googlecode.tesseract.android.TessBaseAPI; //导入依赖的package包/类
/**
 * Initializes Tesseract library using traineddata file.
 * Should not be called directly, is public for testing.
 */
public void init() {
    baseApi = new TessBaseAPI();
    baseApi.setDebug(true);
    String path = Environment.getExternalStorageDirectory() + "/" + Util.FOLDER_DIGITAL_VOTING_PASS + "/";
    File trainedDataFile = new File(path, TesseractOCR.FOLDER_TESSERACT_DATA + "/" + trainedData);
    try {
        mDeviceStorageAccessLock.acquire();
        if (!trainedDataFile.exists()) {
            Log.i(TAG, "No existing trained data found, copying from assets..");
            Util.copyAssetsFile(assetManager.open(trainedData), trainedDataFile);
        } else {
            Log.i(TAG, "Existing trained data found");
        }
        mDeviceStorageAccessLock.release();
        baseApi.init(path, trainedData.replace(TesseractOCR.TRAINED_DATA_EXTENSION, "")); //extract language code from trained data file
    } catch (IOException | InterruptedException e) {
        e.printStackTrace();
        //TODO show error to user, coping failed
    }
}
 

import com.googlecode.tesseract.android.TessBaseAPI; //导入依赖的package包/类
/**
 * Method to start or restart recognition after the OCR engine has been initialized,
 * or after the app regains focus. Sets state related settings and OCR engine parameters,
 * and requests camera initialization.
 */
void resumeOCR() {
    Log.d(TAG, "resumeOCR()");

    // This method is called when Tesseract has already been successfully initialized, so set
    // isEngineReady = true here.
    isEngineReady = true;

    isPaused = false;

    if (handler != null) {
        handler.resetState();
    }
    if (baseApi != null) {
        baseApi.setPageSegMode(pageSegmentationMode);
        baseApi.setVariable(TessBaseAPI.VAR_CHAR_BLACKLIST, characterBlacklist);
        baseApi.setVariable(TessBaseAPI.VAR_CHAR_WHITELIST, characterWhitelist);
    }

    if (hasSurface) {
        // The activity was paused but not stopped, so the surface still exists. Therefore
        // surfaceCreated() won't be called, so init the camera here.
        initCamera(surfaceHolder);
    }
}
 

import com.googlecode.tesseract.android.TessBaseAPI; //导入依赖的package包/类
private List<OcrChar> processOcrIterator(ResultIterator iterator){

        List<OcrChar> ocrChars = new ArrayList<>();

        if (iterator.next(TessBaseAPI.PageIteratorLevel.RIL_SYMBOL)){
            iterator.begin();
        }
        else {
            return ocrChars;
        }

        do {
            List<Pair<String, Double>> choices = iterator.getChoicesAndConfidence(TessBaseAPI.PageIteratorLevel.RIL_SYMBOL);
            int[] pos = iterator.getBoundingBox(TessBaseAPI.PageIteratorLevel.RIL_SYMBOL);
            ocrChars.add(new OcrChar(choices, pos));
        } while (iterator.next(TessBaseAPI.PageIteratorLevel.RIL_SYMBOL));

        iterator.delete();

        return ocrChars;
    }
 

import com.googlecode.tesseract.android.TessBaseAPI; //导入依赖的package包/类
@Override
protected String doInBackground(String... params) {
	runOnUiThread(new Runnable() {
		@Override
		public void run() {
			ProgressWheel pBar = (ProgressWheel) findViewById(R.id.progressBar);
			pBar.setVisibility(View.VISIBLE);
			pBar.spin();
			pBar.setText("Fetching...");		
		}
	});
	// Sets Up Tesseract for text recognition.
	baseApi = new TessBaseAPI();
	baseApi.init(Environment.getExternalStorageDirectory() +   "/tesseract/", "eng");
	//baseApi.setVariable(TessBaseAPI.VAR_CHAR_WHITELIST,"ABCDEFGHIJKLMNOPQRSTUVWXYZ");
	baseApi.setVariable("load_system_dawg","F");
	baseApi.setVariable("load_freq_dawg","F");
	baseApi.setPageSegMode(TessBaseAPI.PageSegMode.PSM_AUTO_OSD);
	for (String key : card.getPatternMap().keySet()) {
		if(card.getPatternMap().get(key).getResource().matches("[t][e][x][t].*")) fetchTextNew(card.getPattern(key).getResource());
	}
	baseApi.end();
	return "sucess";
}
 

import com.googlecode.tesseract.android.TessBaseAPI; //导入依赖的package包/类
public OCRFrameWorker(OnCameraPreviewListener.FrameDrawer drawer) {
    super(drawer);

    this.tessBaseAPI = new TessBaseAPI();
    this.tessBaseAPI.setPageSegMode(PageSegMode.PSM_AUTO_OSD);

    this.tessBaseAPI.setVariable(TessBaseAPI.VAR_ACCURACYVSPEED,
            String.valueOf(50));
    this.tessBaseAPI.init(Environment.getExternalStorageDirectory().getPath(), "eng");

    paint = new Paint();
    paint.setColor(Color.RED);
    int size = (int) TypedValue.applyDimension(TypedValue.COMPLEX_UNIT_DIP,
            (float) 11.0, drawer.getDisplayMetrics());
    paint.setTextSize(size);

    bounds = new Rect();
    paint.getTextBounds("Q", 0, 1, bounds);

    lines = drawer.getDisplayMetrics().heightPixels / bounds.height();

}
 

import com.googlecode.tesseract.android.TessBaseAPI; //导入依赖的package包/类
@Override
protected void onActivityResult(int requestCode, int resultCode, Intent data) {
    if (resultCode != RESULT_OK) {
        return;
    }

    if (requestCode == cameraResultCode) {
        Bundle extras = data.getExtras();
        Bitmap bmp = (Bitmap) extras.get("data");
        image.setImageBitmap(bmp);
        
        TessBaseAPI baseApi = new TessBaseAPI();
        baseApi.init("/mnt/sdcard/tesseract/", "chi_tra");
        baseApi.setPageSegMode(TessBaseAPI.PageSegMode.PSM_SINGLE_LINE);
        baseApi.setImage(bmp);
        String outputText = baseApi.getUTF8Text();
        tessResults.setText(outputText);
    }


    super.onActivityResult(requestCode, resultCode, data);
}
 

import com.googlecode.tesseract.android.TessBaseAPI; //导入依赖的package包/类
@Override
protected void onCreate(Bundle savedInstanceState) {
    super.onCreate(savedInstanceState);
    setContentView(R.layout.activity_orc_bitmap);
    findView();

    mTextBitmap = getTextImage("你好", 300, 300);
    image.setImageBitmap(mTextBitmap);

    baseApi = new TessBaseAPI();
    baseApi.init("/mnt/sdcard/tesseract/", "chi_tra");
    baseApi.setPageSegMode(TessBaseAPI.PageSegMode.PSM_SINGLE_LINE);
    baseApi.setImage(mTextBitmap);
    String outputText = baseApi.getUTF8Text();
    tessResults.setText(outputText);
}
 

import com.googlecode.tesseract.android.TessBaseAPI; //导入依赖的package包/类
/** 
 * Method to start or restart recognition after the OCR engine has been initialized,
 * or after the app regains focus. Sets state related settings and OCR engine parameters,
 * and requests camera initialization.
 */
void resumeOCR() {
  Log.d(TAG, "resumeOCR()");
  
  // This method is called when Tesseract has already been successfully initialized, so set 
  // isEngineReady = true here.
  isEngineReady = true;
  
  isPaused = false;

  if (handler != null) {
    handler.resetState();
  }
  if (baseApi != null) {
    baseApi.setPageSegMode(pageSegmentationMode);
    //baseApi.setVariable(TessBaseAPI.VAR_CHAR_BLACKLIST, characterBlacklist);
    baseApi.setVariable(TessBaseAPI.VAR_CHAR_WHITELIST, "ABCDEFabcdef0123456789");
  }

  if (hasSurface) {
    // The activity was paused but not stopped, so the surface still exists. Therefore
    // surfaceCreated() won't be called, so init the camera here.
    initCamera(surfaceHolder);
  }
}
 

import com.googlecode.tesseract.android.TessBaseAPI; //导入依赖的package包/类
@SmallTest
public void testSetImage() {
    // First, make sure the eng.traineddata file exists.
    assertTrue("Make sure that you've copied " + DEFAULT_LANGUAGE + ".traineddata to "
            + EXPECTED_FILE, new File(EXPECTED_FILE).exists());

    // Attempt to initialize the API.
    final TessBaseAPI baseApi = new TessBaseAPI();
    baseApi.init(TESSBASE_PATH, DEFAULT_LANGUAGE);

    // Set the image to a Bitmap.
    final Bitmap bmp = Bitmap.createBitmap(640, 480, Bitmap.Config.ARGB_8888);
    baseApi.setImage(bmp);

    // Attempt to stesthut down the API.
    baseApi.end();
}
 

import com.googlecode.tesseract.android.TessBaseAPI; //导入依赖的package包/类
/**
 * Method to start or restart recognition after the OCR engine has been
 * initialized, or after the app regains focus. Sets state related settings
 * and OCR engine parameters, and requests camera initialization.
 */
public void resumeOcrEngine() {
    Log.d(TAG, "resumeOcrEngine()");

    // This method is called when Tesseract has already been successfully
    // initialized, so set
    // isEngineReady = true here.

    if (mBaseApi != null) {
        if (mCaptureActivityHandler != null) {
            mCaptureActivityHandler.startDecode(mBaseApi);
        }

        mBaseApi.setPageSegMode(PAGE_SEGMENTATION_MODE);
        mBaseApi.setVariable(TessBaseAPI.VAR_CHAR_BLACKLIST, "");
        mBaseApi.setVariable(TessBaseAPI.VAR_CHAR_WHITELIST, CHARACTER_WHITELIST);
    }
}
 

import com.googlecode.tesseract.android.TessBaseAPI; //导入依赖的package包/类
/**
 * Requests initialization of the OCR engine with the given parameters.
 *
 * @param storageRoot  Path to location of the tessdata directory to use
 * @param languageCode Three-letter ISO 639-3 language code for OCR
 * @param languageName Name of the language for OCR, for example, "English"
 */
private void initOcrEngine(File storageRoot, String languageCode, String languageName) {
    // Set up the dialog box for the thermometer-style download progress
    // indicator
    if (mInitOcrProgressDialog != null) {
        mInitOcrProgressDialog.dismiss();
    }
    mInitOcrProgressDialog = new ProgressDialog(this);

    // Start AsyncTask to install language data and init OCR
    new OcrInitAsyncTask(this,
            new TessBaseAPI(),
            mInitOcrProgressDialog,
            languageCode,
            languageName,
            OCR_ENGINE_MODE).execute(storageRoot.toString());
}
 

import com.googlecode.tesseract.android.TessBaseAPI; //导入依赖的package包/类
/**
 * Returns a string that represents which OCR engine(s) are currently set to be run.
 *
 * @return OCR engine mode
 */
String getOcrEngineModeName() {
    String ocrEngineModeName = "";
    String[] ocrEngineModes = getResources().getStringArray(R.array.ocrenginemodes);
    if (ocrEngineMode == TessBaseAPI.OEM_TESSERACT_ONLY) {
        ocrEngineModeName = ocrEngineModes[0];
    } else if (ocrEngineMode == TessBaseAPI.OEM_CUBE_ONLY) {
        ocrEngineModeName = ocrEngineModes[1];
    } else if (ocrEngineMode == TessBaseAPI.OEM_TESSERACT_CUBE_COMBINED) {
        ocrEngineModeName = ocrEngineModes[2];
    }
    return ocrEngineModeName;
}
 

import com.googlecode.tesseract.android.TessBaseAPI; //导入依赖的package包/类
OcrRecognizeAsyncTask(CaptureActivity activity, TessBaseAPI baseApi, byte[] data, int width, int height) {
  this.activity = activity;
  this.baseApi = baseApi;
  this.data = data;
  this.width = width;
  this.height = height;
}
 

import com.googlecode.tesseract.android.TessBaseAPI; //导入依赖的package包/类
@Override
public void run() {
    if (originBitmap == null) {
        scanFailed(getString(R.string.string_null_pic_single));
        return;
    }
    if (originBitmap.getConfig() != Bitmap.Config.ARGB_8888) {
        scanFailed(getString(R.string.string_pic_not_a888));
        return;
    }
    try {
        TessBaseAPI api = new TessBaseAPI();
        api.init(Common.getTessDataDir(context), Common.getLanguage(context));
        api.setImage(originBitmap.copy(Bitmap.Config.ARGB_8888, true));
        final String result = api.getUTF8Text();
        runOnUiThread(new Runnable() {
            @Override
            public void run() {
                if (isStarting && !isFinishing()) {
                    scanSuccess(result);
                }
            }
        });
        originBitmap.recycle();
        api.clear();
        api.end();
    } catch (final Exception e) {
        e.printStackTrace();
        runOnUiThread(new Runnable() {
            @Override
            public void run() {
                if (getString(R.string.error_tessdata_folder).equals(e.getMessage())) {
                    scanFailed(getString(R.string.string_tess_not_ready));
                } else {
                    scanFailed(e.getMessage());
                }
            }
        });
    }
}