I try OCR image by Google Vision API:
Result text return from API:
"text": "ROBERT M. SCHARF. M.D\n1645 Dorchester Drive\nPlano, Texas 75075\nTELEPHONE (972) 508-3328\nDEA REG. #AS 6975342\nJohn Doe\nDATE 6/1/08\nSPHERICAL\nCYLINDRICAL\nAXIS\nPRISM\nBASE\nOD-3.00\nos3.75\nOD+2.50\nos +2.50\n+1.25\n+1.75\nbi or tri\nD.V\n125\nN.V\nREMARKS +2.75 Hoya GP\n"
Result text i get from website of google: https://cloud.google.com/vision/docs/drag-and-drop
"text": "ROBERT M. SCHARF, M.D.\n1645 Dorchester Drive\nPlano, Texas 75075\nTELEPHONE (072) 596-3328\nDEA REG. #AS 6075342\nFOR John Doe\n- DATE_6/1/08\nPRISM\nBASE\nADDRESS\nB\nOD\nDV\nos\nS PHERICAL\n-3.00\n-3.75\n+2.50\n+2.50\nCYLINDRICAL\n+1.25\n+1.75\nbi or tri\nAXIS\n55\n125\nOS\nREMARKS +2.75 Hoya GP\nRobertas charl, mb.\n"
From website of google :
It can get correct value [-3.75
], but from API , it only return [3.75
]
Why result of google vision api difference with ocr from website?
The cause: I using type=TEXT_DETECTION, web using DOCUMENT_TEXT_DETECTION