vendredi 1 juillet 2016

viewing the Gabor filtered images


I am trying apply Gabor Filter for thetha=0,pi/4,pi/2,3pi/4,pi and want to see the resulting 5 images. How to view those 5 images. I cannot clearly understand the code given in the link I have been trying to follow. Here is my code snippet def build_filters(sigma_size,kernel_size): filters = [] ksize = kernel_size for theta in np.arange(0, np.pi, np.pi / 5): params = {'ksize':(ksize, ksize), 'sigma':sigma_size, 'theta':theta, 'lambd':15.0, 'gamma':0.02, 'psi':0, 'ktype':cv2.CV_32F} kern = cv2.getGaborKernel(**params) kern /= 1.5*kern.sum() filters.append((kern,params)) return filters def process(image,filters): results = [] for kern, params in filters: fimg = cv2.filter2D(image, cv2.CV_8UC3, kern) results.append([fimg]) return results def Gabor_features(image,filters): filtered_images = process(image,filters) return filtered_images I am trying to call this function in main.py imported above file as FE gaborfilters = GaborFilterBank.build_filters(2.0,39) # matlab sig=2.0,kernel size 39 feature_2 = FE.Gabor_features(grayImage,gaborfilters) # here I want to view the 5 filtered images fig, ax = plt.subplots(nrows=1, ncols=5) ax[0].imshow(??) ax[1].imshow(??) ax[2].imshow(??) ax[3].imshow(??) ax[4].imshow(??) Does the 'results.append' is making a structure of all 5 images that I have calculated ? if not how ? please suggest. I tried to follow Gabor Filtering openCV

Aucun commentaire:

Enregistrer un commentaire