Test your AI Agents with the all-new Agent to Agent Testing Platform.Learn More

How to use ocrEngine method in fMBT

Best Python code snippet using fMBT_python

ocr.py

Source:ocr.py

1import requests2from io import BytesIO3import base644import hashlib5import time6import json7from .errors import *8class OCREngine(object):9    def ocr(self, img):10        pass11class OCRBox(object):12    def __init__(self, x, y, w, h):13        self.__x = x14        self.__y = y15        self.__w = w16        self.__h = h17    @property18    def left(self):19        return self.__x20    @property21    def right(self):22        return self.__x + self.__w23    @property24    def top(self):25        return self.__y26    @property27    def bottom(self):28        return self.__y + self.__h29    @property30    def width(self):31        return self.__w32    @property33    def height(self):34        return self.__h35    def __repr__(self):36        return 'OCRBox(%d,%d,%d,%d)' % (self.__x, self.__y, self.__w, self.__h)37class OCRWord(OCRBox):38    def __init__(self, location, word, chars):39        self.__word = word40        self.__chars = chars41        super().__init__(location.left, location.top, location.width, location.height)42    @property43    def word(self):44        return self.__word45    @property46    def chars(self):47        return self.__chars48    def __len__(self):49        return len(self.word)50    def __getitem__(self, key):51        boxes = self.chars[key]52        left = min([it.left for it in boxes])53        right = max([it.right for it in boxes])54        top = min([it.top for it in boxes])55        bottom = max([it.bottom for it in boxes])56        return OCRWord(OCRBox(left, top, right - left, bottom - top), self.word[key], boxes)57    def __repr__(self):58        return 'OCRWord(%s,%s,%s)' % (super().__repr__(), self.__word.__repr__(), self.chars.__repr__())59class BaiduOCREngine(OCREngine):60    def __init__(self, token, accurate=False):61        self.__token = token62        self.__url = 'https://aip.baidubce.com/rest/2.0/ocr/v1/' + ('accurate' if accurate else 'general')63        self.__headers = {'Content-Type': 'application/x-www-form-urlencoded'}64        self.__params = {'access_token': token}65        self.__body = {'recognize_granularity': 'small'}66    @staticmethod67    def __loc2box(loc):68        return OCRBox(loc['left'], loc['top'], loc['width'], loc['height'])69    def ocr(self, img):70        byte = BytesIO()71        img.save(byte, format='png')72        encoded = base64.b64encode(byte.getvalue()).decode()73        self.__body['image'] = encoded74        response = requests.post(url=self.__url, params=self.__params, headers=self.__headers, data=self.__body).json()75        if 'code' in response:76            raise BaiduOCRError(code=response['code'], message=response['message'])77        results = response['words_result']78        words = [79            OCRWord(80                self.__loc2box(word['location']),81                ''.join([c['char'] for c in word['chars']]),82                [self.__loc2box(c['location']) for c in word['chars']]83            ) for word in results84        ]85        return words86class SogouOCREngine(OCREngine):87    def __init__(self, pid, key):88        self.__pid = pid89        self.__key = key90        self.__service = 'basicOpenOcr'91        self.__url = 'http://deepi.sogou.com/api/sogouService'92        self.__headers = {'content-type': "application/x-www-form-urlencoded", 'accept': "application/json"}93        self.__params = {'pid': pid, 'key': key, 'service': self.__service}94        self.__body = {}95    @staticmethod96    def __frame2box(frame):97        x0, y0 = map(int, frame[0].split(','))98        x1, y1 = map(int, frame[1].split(','))99        x2, y2 = map(int, frame[2].split(','))100        x3, y3 = map(int, frame[3].split(','))101        left = min(x0, x1, x2, x3)102        right = max(x0, x1, x2, x3)103        top = min(y0, y1, y2, y3)104        bottom = max(y0, y1, y2, y3)105        return OCRBox(left, top, right - left, bottom - top)106    @staticmethod107    def __split(box, n):108        det = box.width / n109        return [OCRBox(round(box.left + det * c), box.top, round(det), box.height) for c in range(n)]110    def ocr(self, img):111        byte = BytesIO()112        img.save(byte, format='png')113        encoded = base64.b64encode(byte.getvalue()).decode()114        self.__body['image'] = encoded115        salt = str(time.time_ns())116        sign = self.__md5(self.__pid + self.__service + salt + encoded[:1024] + self.__key)117        self.__params['salt'] = salt118        self.__params['sign'] = sign119        response = requests.post(url=self.__url, params=self.__params, headers=self.__headers, data=self.__body).json()120        if 'code' in response:121            raise SogouOCRError(code=response['code'], message=response['message'])122        if 'err_code' in response:123            raise SogouOCRError(code=response['err_code'], message=response['err_msg'])124        results = response['result']125        words = [126            OCRWord(127                self.__frame2box(word['frame']),128                word['content'].strip(),129                self.__split(self.__frame2box(word['frame']), len(word['content'].strip()))130            ) for word in results131        ]132        return words133    @staticmethod134    def __md5(s):135        return hashlib.md5(s.encode('utf8')).hexdigest()136class AliOCREngine(OCREngine):137    def __init__(self, appcode):138        self.__appcode = appcode139        self.__url = 'https://ocrapi-advanced.taobao.com/ocrservice/advanced'140        self.__headers = {'content-type': 'application/json', 'Authorization': 'APPCODE ' + appcode}141        self.__body = {'charInfo': True}142    @staticmethod143    def __pos2box(pos):144        left = pos[0]['x']145        top = pos[0]['y']146        right = pos[2]['x']147        bottom = pos[2]['y']148        return OCRBox(left, top, right - left, bottom - top)149    @staticmethod150    def __char2box(char):151        return OCRBox(char['x'], char['y'], char['w'], char['h'])152    def ocr(self, img):153        byte = BytesIO()154        img.save(byte, format='png')155        encoded = base64.b64encode(byte.getvalue()).decode()156        self.__body['img'] = encoded157        response = requests.post(url=self.__url, headers=self.__headers, data=json.dumps(self.__body)).json()158        if 'error_code' in response:159            raise AliOCRError(code=response['error_code'], message=response['error_msg'])160        results = response['prism_wordsInfo']161        words = [162            OCRWord(163                self.__pos2box(word['pos']),164                ''.join([c['word'] for c in word['charInfo']]),165                [self.__char2box(c) for c in word['charInfo']]166            ) for word in results167        ]...

ocrdiffer.py

Source:ocrdiffer.py

1from .ocr import *2from .match import match_box3from .diff import diff_words4from .draw import highlight5from .split import split_word6from io import BytesIO7from PIL import Image8import base649class OCRDiffer(object):10    def __init__(self, engine='baidu', *args, **kwargs):11        if engine == 'baidu':12            self.__engine = BaiduOCREngine(*args, **kwargs)13        elif engine == 'sogou':14            self.__engine = SogouOCREngine(*args, **kwargs)15        elif engine == 'ali':16            self.__engine = AliOCREngine(*args, **kwargs)17        else:18            raise OCREngineUnknownError()19    def work(self, img1, img2):20        img2.resize(img1.size)21        ocr1 = self.__engine.ocr(img1)22        ocr2 = self.__engine.ocr(img2)23        ocr1 = split_word(ocr1)24        ocr2 = split_word(ocr2)25        mat, unmat1, unmat2 = match_box(ocr1, ocr2)26        diff1, diff2 = diff_words(mat)27        out = highlight(img1, img2, ocr1, ocr2, diff1 + unmat1, diff2 + unmat2)28        return out29    def work_http(self, img1, img2):30        dec1 = base64.b64decode(img1)31        dec2 = base64.b64decode(img2)32        img1 = Image.open(BytesIO(dec1))33        img2 = Image.open(BytesIO(dec2))...

letssee.py

Source:letssee.py

1from asprise_ocr_api import *2Ocr.set_up()3ocrEngine=Ocr()4ocrEngine.start_engine("eng")5s=ocrEngine.recognize("XAO.png",-1,-1,-1,-1,-1,OCR_RECOGNIZE_TYPE_ALL,OCR_OUTPUT_FORMAT_PLAINTEXT)6print("Result ->"+s)...

Automation Testing Tutorials

Learn to execute automation testing from scratch with LambdaTest Learning Hub. Right from setting up the prerequisites to run your first automation test, to following best practices and diving deeper into advanced test scenarios. LambdaTest Learning Hubs compile a list of step-by-step guides to help you be proficient with different test automation frameworks i.e. Selenium, Cypress, TestNG etc.