sim_train kendim yazdım.
svm_train:
# coding=utf-8_
import cv2
import numpy as np
from numpy.linalg import norm
svm_params = dict(kernel_type=cv2.ml.SVM_RBF, svm_type=cv2.ml.SVM_C_SVC,C=2.67, gamma=5.383)
class StatModel(object):
def load(self, fn):
self.model.load(fn) # python rapper bug
def save(self, fn):
self.model.save(fn)
class SVM(StatModel):
def __init__(self, C=1, gamma=0.5):
#self.model = cv2.SVM()
self.model = cv2.ml.SVM_create()
self.model.setGamma(gamma)
self.model.setC(C)
self.model.setKernel(cv2.ml.SVM_RBF)
self.model.setType(cv2.ml.SVM_C_SVC)
def train(self, samples, responses):
#self.model.train(samples, responses,params=svm_params) # inbuilt training function
self.model.setType(cv2.ml.SVM_C_SVC)
self.model.setC(1)
self.model.setKernel(cv2.ml.SVM_RBF)
self.model.setGamma(.1)
self.model.train(samples, cv2.ml.ROW_SAMPLE, responses.astype(int))
def predict(self, samples):
#return self.model.predict_all(samples).ravel()
_ret, resp = self.model.predict(samples)
return resp.ravel()
def preprocess_hog(digits):
samples = []
for img in digits:
gx = cv2.Sobel(img, cv2.CV_32F,1,0) #Resmin X ekseninine gore kenarliklarini belirtir.
gy = cv2.Sobel(img, cv2.CV_32F,0,1) #Resmin Y ekseninine gore kenarliklarini belirtir.
mag, ang = cv2.cartToPolar(gx, gy) #2D vektörlerin büyüklüğünü ve açısını hesaplar.
bin_n = 16
bin = np.int32(bin_n * ang / (2 * np.pi))
bin_cells = bin[:100, :100], bin[100:, :100], bin[:100, 100:], bin[100:, 100:]
mag_cells = mag[:100, :100], mag[100:, :100], mag[:100, 100:], mag[100:, 100:]
hists = [np.bincount(b.ravel(), m.ravel(), bin_n) for b, m in zip(bin_cells, mag_cells)]
hist = np.hstack(hists)
# transform to Hellinger kernel
eps = 1e-7
hist /= hist.sum() + eps
hist = np.sqrt(hist)
hist /= norm(hist) + eps
samples.append(hist)
return np.float32(samples)
# Here goes my wrappers:_
def hog_single(img):
samples = []
gx = cv2.Sobel(img, cv2.CV_32F, 1, 0)
gy = cv2.Sobel(img, cv2.CV_32F, 0, 1)
mag, ang = cv2.cartToPolar(gx, gy)
bin_n = 16
bin = np.int32(bin_n * ang / (2 * np.pi))
bin_cells = bin[:100, :100], bin[100:, :100], bin[:100, 100:], bin[100:, 100:]
mag_cells = mag[:100, :100], mag[100:, :100], mag[:100, 100:], mag[100:, 100:]
hists = [np.bincount(b.ravel(), m.ravel(), bin_n) for b, m in zip(bin_cells, mag_cells)]
hist = np.hstack(hists)
# transform to Hellinger kernel
eps = 1e-7
hist /= hist.sum() + eps
hist = np.sqrt(hist)
hist /= norm(hist) + eps
samples.append(hist)
return np.float32(samples)
def trainSVM():
imgs = []
num=17
for i in range(65, num + 65):
for j in range(1, 401):
print(
'Class ' + chr(i) + ' is being loaded ')
imgs.append(cv2.imread('DataSet/' + chr(i) + '_' + str(j) + '.jpg', 0)) # all images saved in a list
labels = np.repeat(np.arange(1, num + 1), 400) # label for each corresponding image saved above
samples = preprocess_hog(imgs) # images sent for pre processeing using hog which returns features for the images
print("SVM is building wait some time ...")
print (len(labels))
print (len(samples))
model=SVM(C=2.67, gamma=5.383)
model.train(samples, labels) # features trained against the labels using svm
return model
def predict(model, img):
samples = hog_single(img)
resp = model.predict(samples)
return resp