Question

当我在python和opencv中计算两个灰度图像之间的相似度时，错误报告了compare_ssim()函数。

gray1 = cv2.cvtColor(frame1, cv2.COLOR_BGR2GRAY)
gray2 = cv2.cvtColor(frame2, cv2.COLOR_BGR2GRAY)
(score, diff) = compare_ssim(gray1, gray2, full=True)

错误消息如下：

TypeError：structural_similarity（）接受2个位置参数，但给出了8个

这是库中的函数。来自_structural_similarity.py文件。目录为\skimage\measure\_structural_similarity.py。

def compare_ssim(X, Y, win_size=None, gradient=False,
                 data_range=None, multichannel=False, gaussian_weights=False,
                 full=False, **kwargs):
    warn('DEPRECATED: skimage.measure.compare_ssim has been moved to '
         'skimage.metrics.structural_similarity. It will be removed from '
         'skimage.measure in version 0.18.', stacklevel=2)
       return structural_similarity(X, Y, win_size, gradient,
                                 data_range, multichannel, gaussian_weights, 
                                 full, **kwargs)

return行，此行中显示错误消息。

和structural_similarity()。此函数的目录为\skimage\metrics\_structural_similarity.py，该函数如下：

def structural_similarity(im1, im2,
                          *,
                          win_size=None, gradient=False, data_range=None,
                          multichannel=False, gaussian_weights=False,
                          full=False, **kwargs):

出什么问题了，为什么？

Answer 1

from skimage.metrics import structural_similarity

尝试...

Answer 2

我尝试使用这个：

from skimage.metrics import structural_similarity as ssim
...
s = ssim(gray_image_1,gray_image_2)

此处，gray_image_1.shape = gray_image_2.shape =（420,420）。这对我有用（python 3.6.6）。

`

Answer 3

感谢您的耐心配合。我已经解决了这个问题。应该是函数调用错误。当我将头文件'from skimage.measures import compare_ssim'替换为from skimage.metrics import structural_similarity，并将函数调用替换为(score, diff) = structural_similarity(gray1, gray2, full=True)时，编译成功

TypeError：structural_similarity（）接受2个位置参数，但给出了8个

3 个答案: