ocr_yogurt.hdev识别酸奶盒保质期 相关例程学习

Index:.../Applications/OCR/ocr_yogurt.hdev


* This example demonstrates how to segment the expiration date of
* a yoghurt cap. After the segmentation, the characters are classified
* by the mlp ocr classifier using the dotprint font.


* Display initializations
dev_update_off ()
dev_close_window ()
read_image (Image, 'ocr/yogurt_lid_01')
dev_open_window_fit_image (Image, 0, 0, -1, -1, WindowHandle)
set_display_font (WindowHandle, 16, 'mono', 'true', 'false')
dev_set_draw ('margin')
dev_set_colored (12)
dev_set_line_width (2)

read_ocr_class_mlp ('DotPrint_0-9A-Z.omc', OCRHandle)
* Approximate the character's dot size
DotDiameter := 6

里面字的dot差不多就是直径为6像素
* Rough character size
CharHeight := 60
CharWidth := 50
* Number of candidates returned by the classifier
NCandidates := 3
for I := 1 to 4 by 1
    read_image (Image, 'ocr/yogurt_lid_' + I$'02')
    * 
    * Segmentation of characters on yoghurt cover
    decompose3 (Image, ImageR, ImageG, ImageB)
    * diff_of_gauss() emphasizes structures having a characteristic dimension of size
    * DotDiameter (line thickness, maximum diameter, ...)
    diff_of_gauss (ImageR, DiffOfGauss, DotDiameter / 2, 2)
    scale_image_max (DiffOfGauss, ImageScaleMax)

这里面把R通道图像做了变换,我感觉好像不做其实也没影响


    * segment_characters expects black on white characters
    invert_image (ImageScaleMax, ImageInvert)
    segment_characters (ImageInvert, ImageInvert, ImageForeground, RegionForeground, 'local_contrast_best', 'false', 'true', 'light', CharWidth, CharHeight, 20, 50, UsedThreshold)

segment_characters 这个基本找到了字符, All segmentation methods assume that the text ist darker than the background. If this is not the case, please invert the image with invert_image. 说明这个算子使用前要把字变黑来找的。
    * Merge dots
    closing_circle (RegionForeground, RegionClosing, 4.0)
    connection (RegionClosing, ConnectedRegions)
    * We assume that characters are aligned along a line and
    * the regions over or below this line are distractors
    select_shape (ConnectedRegions, SelectedRegions, ['height','width','area'], 'and', [CharHeight - 20,CharWidth - 20,210], [CharHeight + 20,CharWidth + 20,CharHeight * CharHeight])

按照字符大小做第一次选择
    area_center (SelectedRegions, Area, Row, Column)
    get_image_size (Image, Width, Height)
    tuple_histo_range (Row, 0, Height, 2 * Height / CharHeight, Histo, BinSize)
    tuple_find (Histo, max(Histo), IndMax)
    CharRow := BinSize * (IndMax[0] + 0.5)

    *把row分成12bins然后histo就有12个,在这里面看分布情况,则5个字符是5,上面一个字符是1
    *则5那个位置的row就是保质期所在的位置了
    select_shape (SelectedRegions, Characters, 'row', 'and', CharRow - CharHeight / 2, CharRow + CharHeight / 2)

然后按照这个选择就可以每次选出来字符的位置


    * 
    * Read out the characters on the yoghurt cover
    sort_region (Characters, SortedRegions, 'character', 'true', 'row')
    select_obj (SortedRegions, Chars, [1,2,3,4,5])
    do_ocr_word_mlp (Chars, ImageForeground, OCRHandle, '\\d{4}[A-Z]', NCandidates, 2, Class, Confidence, Word, Score)

最后这个识别,参数什么的,照着例程调就行。
    * 
    DateString := Word{0} + Word{1} + '.' + Word{2} + Word{3} + '.'
    BatchID := Word{4}
    * 
    dev_display (Image)
    dev_display (Chars)
    disp_message (WindowHandle, 'Best before: ' + DateString + '\nBatch ID   : ' + BatchID, 'window', 12, 12, 'black', 'true')
    if (I < 4)
        disp_continue_message (WindowHandle, 'black', 'true')
        stop ()
    endif
endfor
clear_ocr_class_mlp (OCRHandle)


  • 0
    点赞
  • 1
    收藏
    觉得还不错? 一键收藏
  • 0
    评论

“相关推荐”对你有帮助么?

  • 非常没帮助
  • 没帮助
  • 一般
  • 有帮助
  • 非常有帮助
提交
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值