edge_eval_python icon indicating copy to clipboard operation
edge_eval_python copied to clipboard

A python implementation of edge eval

Results 11 edge_eval_python issues
Sort by recently updated
recently updated
newest added

您好,在使用此代码进行Multicue-boundary数据集评测时,会出现IndexError:tuple index out of range 这个错误,但在BSDS500数据集中可以正常运行,二者GT都是由5人平均得来的,形式相同。 报错位置为 `gt = [g.item()[1] for g in loadmat(gt)["groundTruth"][0]] # 0: Segmentation, 1: Boundaries` 将`g.item()[1]`改成`g.item()[0]`后可运行代码,但结果和Matlab版不同。 请问应当如何解决?感谢您的回复。

Hello, I just came into contact with image segmentation. I don't understand the whole segmentation process very well. How do you preprocess the image? After obtaining the gray value of...

Hello, thank you for translating this matlab eval_function into Python. I have some question about your edge_eval_python, how do I use this function to evaluate my dataset? I found that...

现在最常用的边缘检测评价标准是基于https://github.com/pdollar/edges 的,但是该代码的核心部分correspondPixels是不开源的 请问您在实现这块代码时,是您独立完成的还是有其他参照 方便分享一下吗? 先行谢过!

Hi, I tried to run your code and encountered an error in function edges_eval_img(): _TypeError: 'int' object is not subscriptable._ I looked into it and found this line hard to...

Hello, it gives segmentation fault on the below line: solver.nms(out.ctypes.data_as(c_float_pointer), edge.ctypes.data_as(c_float_pointer), ori.ctypes.data_as(c_float_pointer), r, s, m, w, h) I normalized image to range 0-1, but error still exists. How can I...

您好,我用您的python版本的eval测试了RCF的结果,同时也用matlab版本的eval做了对比,发现结果不一样的—— ![image](https://user-images.githubusercontent.com/24868875/113238476-75655380-92db-11eb-9c60-2e6361f0d4f3.png) 以上是python版本的 ![image](https://user-images.githubusercontent.com/24868875/113238514-83b36f80-92db-11eb-9e06-b3df610f8fb4.png) 这个是matlab版本的 结果有误差这是一个问题,关键是OIS不可能会低于ODS的。 最后还是感谢您的python代码,根据您的代码思路,我也对matlab版本代码有了一定的了解

不是有效的Win32应用程序

In bwmorph_thin.py, the numpy dtype for some of the arrays is dtype=np.bool. I think this is deprecated in recent versions of NumPy. Full error: "AttributeError: module 'numpy' has no attribute...