Image recognition of numbers

Hello

I'm taking a photo of some images and trying to recognize automatically what number i have.

Attaching an example where i coudn't recognize the image even though it's easy to see with the eye.

both str2double(ocrOutput.Text) and ocrOutput.WordConfidences

Are empty.

Any idea how to settle this?

ocrOutput = ocr(TEMP, 'TextLayout', 'Line', 'characterSet', '0123456789')

Answer 1

[Eli]

So I found a solution.

Appereanly the OCR prefers white background and black charecters (and not the other way around)

so the following code does the trick:

load image
imshow(TEMP)

BW = ~imbinarize(TEMP);
imshow(BW)

ocrOutput = ocr(BW, 'TextLayout', 'word', 'characterSet', '0123456789')

ocrOutput = 
  ocrText with properties:

                      Text: '297183↵↵'
    CharacterBoundingBoxes: [8×4 double]
      CharacterConfidences: [8×1 single]
                     Words: {'297183'}
         WordBoundingBoxes: [11 12 638 165]
           WordConfidences: 0.9626
                 TextLines: {'297183'}
     TextLineBoundingBoxes: [11 12 638 165]
       TextLineConfidences: 0.9626

Answer 2

[Voss]

Try changing the specified Layout type to 'word'

load image
imshow(TEMP)

ocrOutput = ocr(TEMP, 'TextLayout', 'word', 'characterSet', '0123456789')

ocrOutput = 
  ocrText with properties:

                      Text: '297183↵↵'
    CharacterBoundingBoxes: [8×4 double]
      CharacterConfidences: [8×1 single]
                     Words: {'297183'}
         WordBoundingBoxes: [11 12 638 165]
           WordConfidences: 0.9594
                 TextLines: {'297183'}
     TextLineBoundingBoxes: [11 12 638 165]
       TextLineConfidences: 0.9594

[Eli]

doesn't work for me for some reason:

ocrOutput =

ocrText with properties:

Text: '8↵↵'

CharacterBoundingBoxes: [3×4 double]

CharacterConfidences: [3×1 single]

Words: {'8'}

WordBoundingBoxes: [11 12 638 165]

WordConfidences: 0

TextLines: {'8'}

TextLineBoundingBoxes: [11 12 638 165]

TextLineConfidences: 0