python数字识别-模型应用

简介

数字识别是计算视觉中一个重要的任务，可以用来完成老设备加装摄像头完成数据采集的工作。本文来介绍下用opencv实现数字识别训练。

代码

import cv2
import numpy as np

class OCR():

    def __init__(self):
        #######   training part    ###############
        samples = np.loadtxt('generalsamples.data',np.float32)
        responses = np.loadtxt('generalresponses.data',np.float32)
        responses = responses.reshape((responses.size,1))
        # model = cv2.KNearest()
        self.model = cv2.ml.KNearest_create()
        # model.train(samples,responses)
        self.model.train(samples, cv2.ml.ROW_SAMPLE, responses)

    def OCR_image(self,image_path):
        ############################# testing part  #########################

        im = cv2.imread(image_path)
        out = np.zeros(im.shape,np.uint8)
        gray = cv2.cvtColor(im,cv2.COLOR_BGR2GRAY)
        thresh = cv2.adaptiveThreshold(gray,255,1,1,11,2)
        contours,hierarchy = cv2.findContours(thresh,cv2.RETR_EXTERNAL,cv2.CHAIN_APPROX_SIMPLE)
        # print(contours)
        # print(hierarchy)
        list = []
        for cnt in contours:
            # if cv2.contourArea(cnt)>5: #面积
                    [x,y,w,h] = cv2.boundingRect(cnt)
            #     # print([x,y,w,h])
            #     if  (h>0 ):
                    # print([x, y, w, h])
                    cv2.rectangle(im,(x,y),(x+w,y+h),(0,255,0),2)
                    roi = thresh[y:y+h,x:x+w]
                    roismall = cv2.resize(roi,(10,10))
                    roismall = roismall.reshape((1,100))
                    roismall = np.float32(roismall)
                    retval, results, neigh_resp, dists = self.model.findNearest(roismall, k = 1)
                    string = str(int((results[0][0])))
                    # print(type(string))
                    # print(chr(int(string)))
                    list.append([int(x),chr(int(string))])
                    cv2.putText(out,chr(int(string)),(x,y+h),0,1,(0,255,0))

        list = sorted(list) #排序
        str1 = ''.join(str(i[1]) for i in list)

        # print("结果：",str1)

        cv2.imshow('im', im)
        cv2.imshow('out', out)
        cv2.waitKey(0)
        return str1

image_OCR = OCR()
print(image_OCR.OCR_image('test1.png'))
# num = np.asarray(list,dtype="int64")
# # data = num[num[:,0].argsort()] #通过x排序
# data = num[np.argsort(num[:,0])]
# # data = data[:,data[2].argsort()]
# print(data)
# data = data[:,1]
# list = data.tolist()
# string = ''.join([str(i) for i in list])
# print(string) #识别后的字符


1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67

执行程序，结果如图，会自动识别不同颜色，不同大小，不同位置的数字。