python opencv ocr_OpenCV-Python教程:53.使用SVM进行手写数据的OCR

import cv2

import numpy as np

SZ=20

bin_n = 16 # Number of bins

svm_params = dict( kernel_type = cv2.SVM_LINEAR, svm_type = cv2.SVM_C_SVC, C=2.67, gamma=5.383 )

affine_flags = cv2.WARP_INVERSE_MAP|cv2.INTER_LINEAR

def deskew(img):

m = cv2.moments(img)

if abs(m['mu02']) < 1e-2:

return img.copy()

skew = m['mu11']/m['mu02']

M = np.float32([[1, skew, -0.5*SZ*skew], [0, 1, 0]])

img = cv2.warpAffine(img,M,(SZ, SZ),flags=affine_flags)

return img

def hog(img):

gx = cv2.Sobel(img, cv2.CV_32F, 1, 0)

gy = cv2.Sobel(img, cv2.CV_32F, 0, 1)

mag, ang = cv2.cartToPolar(gx, gy)

bins = np.int32(bin_n*ang/(2*np.pi))    # quantizing binvalues in (0...16)

bin_cells = bins[:10,:10], bins[10:,:10], bins[:10,10:], bins[10:,10:]

mag_cells = mag[:10,:10], mag[10:,:10], mag[:10,10:], mag[10:,10:]

hists = [np.bincount(b.ravel(), m.ravel(), bin_n) for b, m in zip(bin_cells, mag_cells)]

hist = np.hstack(hists)    # hist is a 64 bit vector

return hist

img = cv2.imread('digits.png',0)

cells = [np.hsplit(row,100) for row in np.vsplit(img,50)]

# First half is trainData, remaining is testData

train_cells = [ i[:50] for i in cells ]

test_cells = [ i[50:] for i in cells]

######    Now training      ########################

deskewed = [map(deskew,row) for row in train_cells]

hogdata = [map(hog,row) for row in deskewed]

trainData = np.float32(hogdata).reshape(-1,64)

responses = np.float32(np.repeat(np.arange(10),250)[:,np.newaxis])

svm = cv2.SVM()

svm.train(trainData,responses, params=svm_params)

svm.save('svm_data.dat')

######    Now testing      ########################

deskewed = [map(deskew,row) for row in test_cells]

hogdata = [map(hog,row) for row in deskewed]

testData = np.float32(hogdata).reshape(-1,bin_n*4)

result = svm.predict_all(testData)

#######  Check Accuracy  ########################

mask = result==responses

correct = np.count_nonzero(mask)

print correct*100.0/result.size