[英]Tesseract - Supposedly easy image turn out to wrong numbers
Please find below some images that tesseract recognized it incorrectly. 请在下面找到一些tesseract无法正确识别的图像。
47 is recognized as "4]". 47被识别为“ 4]”。
55 is recognized as "S55". 55被识别为“ S55”。
90 is recognized as "IQ". 90被识别为“ IQ”。
I thought the images are pretty good and should be easy to be recognized by Tesseract. 我认为这些图像非常好,应该容易被Tesseract识别。 But the results turn out to be wrong.
但是结果证明是错误的。 The code I used is shown below.
我使用的代码如下所示。
import cv2
import pytesseract
from PIL import Image
import glob
for i in glob.glob('*.png'):
img = cv2.imread(i, 0)
tessdata_dir_config = '--tessdata-dir "C:\Program Files (x86)\Tesseract-OCR\" --psm 10'
result = pytesseract.image_to_string(Image.fromarray(img), config=tessdata_dir_config)
print result
Does anyone know what is going on and how to improve the performance? 有谁知道发生了什么事以及如何提高性能?
Okay, I find an answer for my question. 好吧,我找到了我的问题的答案。 It seems that Tesseract doesn't like bold characters, so you have to erode the black part of the characters a little bit.
Tesseract似乎不喜欢粗体字符,因此您必须稍微侵蚀字符的黑色部分。 But beware of that
cv2.erode
will erode white part of the characters, so we have to use cv2.dilate
to achieve the objective. 但是请注意,
cv2.erode
将侵蚀字符的白色部分,因此我们必须使用cv2.dilate
来达到目的。
for i in ['47-4].png', '55-S55.png', '90-IQ.png']:
img = cv2.imread(i, 0)
### After apply dilation using 3X3 kernal. The recognition results are improved.##
kernel = np.ones((3, 3), np.uint8)
img = cv2.dilate(img, kernel, iterations=2)
cv2.imwrite("./output/" + i[:-4]+'_dilate.png', img)
tessdata_dir_config = '--tessdata-dir "D:\Program Files\Tesseract-ocr\" --psm 10'
result = pytesseract.image_to_string(Image.fromarray(img), config=tessdata_dir_config)
print result
I would like to see if there are any better analysis to this question. 我想看看这个问题是否有更好的分析。 So I would let it open for a while and choose the best answer.
因此,我将其打开一段时间并选择最佳答案。
I had the problem of reading text from android device screens. 我有从Android设备屏幕读取文本的问题。 On some devices it worked on others didn't.
在某些设备上,它无法在其他设备上运行。 I found in tesseract documentation that it has something to do with image dpi.
我在tesseract 文档中发现,它与图像dpi有关。
Tesseract works best on images which have a DPI of at least 300 dpi, so it may be beneficial to resize images.
Tesseract在DPI至少为300 dpi的图像上效果最佳,因此调整图像尺寸可能会有所帮助。 For more information see the FAQ.
有关更多信息,请参见FAQ。
So I used resize function of cv2 to rescale the image. 因此,我使用了cv2的大小调整功能来重新缩放图像。
path = "/home/share/workspace/NNW4JJ4T4LR4G66H_ZTE_Blade_L5/clock_present_cropped.png"
path2 = "/home/share/workspace/NNW4JJ4T4LR4G66H_ZTE_Blade_L5/clock_present_cropped_2.png"
crop_img2 = cv2.imread(str(path))
img_scaled = cv2.resize(crop_img2, None, fx=0.5, fy=0.5, interpolation=cv2.INTER_LINEAR)
cv2.imwrite(str(path2), img_scaled)
crop_img2 = Image.open(path2)
result = pytesseract.image_to_string(crop_img2)
Now it works well with all devices. 现在,它适用于所有设备。
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.