` I have the following code :
import cv2
import os
from os import listdir
import numpy as np
from PIL import Image
from tabulate import tabulate
import itertools
#sift
sift = cv2.SIFT_create()
#feature matching
bf = cv2.BFMatcher(cv2.NORM_L2, crossCheck=True)
# get the path/directory
folder_dir = "./runs/myDetect/SIFT"
col_names = []
data = []
all_keypoints = []
all_descriptors = []
for image in os.listdir(folder_dir):
# check if the image ends with png or jpg or jpeg
if (image.endswith(".png") or image.endswith(".jpg") or image.endswith(".jpeg")):
col_names = ["KeyPoints lenght", "Numbers", "Keypoints"]
opened_img = np.array(Image.open(folder_dir image))
gray_img= cv2.cvtColor(opened_img, cv2.COLOR_BGR2GRAY)
keypoints, descriptors = sift.detectAndCompute(gray_img, None)
data.append([image, len(keypoints)])
all_keypoints.append([image, keypoints])
all_descriptors.append([descriptors]) #all_descriptors.append([image, descriptors])
print(tabulate(data, headers=col_names, tablefmt="fancy_grid"))
for a, b in itertools.combinations(all_descriptors, 2):
a=np.array(a).astype('uint8')
print(type(a))
b=np.array(b).astype('uint8')
print(type(b))
if type(a)!=type(None) and type(b)!=type(None) :
if a or b is None:
print(False)
matches = bf.match(a,b)
matches = sorted(matches, key = lambda x:x.distance)
cv2.waitKey(1)
cv2.destroyAllWindows()
I am trying to find similarities between 2 images on the same file. I am using SIFT. Output of SIFT are keypoints and descriptor. I created a list named all_descriptors and for each image I add new descriptor to this list. Finally, I want to compare this descriptors between each other. On this part matches = bf.match(a,b)
of the code, I receive following error : `matches = bf.match(a,b)
cv2.error: OpenCV(4.6.0) /io/opencv/modules/core/src/batch_distance.cpp:274: error: (-215:Assertion failed) type == src2.type() && src1.cols == src2.cols && (type == CV_32F || type == CV_8U) in function 'batchDistance'. What is the solution? How can I compare 2 images in the same file?
CodePudding user response:
Please try the same code with steps in below
- all_descriptors.append(descriptors)
- a=np.array(a).astype('float32')