Development of the ocr part of AOI
Samo Penic
2018-11-16 e2fa6a35a6548a5acfd000af56ea333df8149b27
Ocr.py
@@ -6,13 +6,15 @@
class Paper:
    def __init__(self, filename=None, sid_classifier=None):
    def __init__(self, filename=None, sid_classifier=None, settings=None):
        self.filename = filename
        self.invalid = None
        self.QRData = None
        self.settings={'answer_treshold':0.25,} if settings is None else settings
        self.errors = []
        self.warnings = []
        self.sid_classifier=sid_classifier
        self.sid=None
        self.sid_classifier = sid_classifier
        if filename is not None:
            self.loadImage(filename)
            self.runOcr()
@@ -216,11 +218,50 @@
    def get_enhanced_sid(self):
        if self.sid_classifier is None:
            return "x"
        es = getSID(
        if self.settings is not None:
            sid_mask=self.settings.get("sid_mask", None)
        es,err,warn = getSID(
            self.img[
                int(0.045 * self.imgHeight) : int(0.085 * self.imgHeight),
                int(0.7 * self.imgWidth) : int(0.99 * self.imgWidth),
            ],
            self.sid_classifier,
            sid_mask
        )
        [self.errors.append(e) for e in err]
        [self.warnings.append(w) for w in warn]
        return es
    def get_code_data(self):
        qrdata = bytes.decode(self.QRData, 'utf8')
        if self.QRDecode[0].type=='EAN13':
            return {'exam_id': int(qrdata[0:7]),
                    'page_no': int(qrdata[7]),
                    'paper_id': int(qrdata[-5:-1]),
                    'faculty_id': None,
                    'sid': None
                    }
        else:
            data=qrdata.split(',')
            retval={'exam_id': int(data[1]),
                    'page_no': int(data[3]),
                    'paper_id':int(data[2]),
                    'faculty_id':int(data[0]),
            }
            if(len(data)>4):
                retval['sid']=data[4]
            return retval
    def get_paper_ocr_data(self):
        data=self.get_code_data()
        data['qr']=self.QRData
        data['errors']=self.errors
        data['warnings']=self.warnings
        data['up_position']=(list(self.xMarkerLocations[1]/self.imgWidth), list(self.yMarkerLocations[1]/self.imgHeight))
        data['right_position']=(list(self.xMarkerLocations[1]/self.imgWidth), list(self.yMarkerLocations[1]/self.imgHeight))
        data['ans_matrix']=((np.array(self.answerMatrix)>self.settings['answer_treshold'])*1).tolist()
        if data['sid'] is None:
            data['sid']=self.get_enhanced_sid()
        return data