walstruzz
walstruzz
assume that you have image `{images_dir}/000012.jpg` and gt `{gt_dir}/000012/xxx.png` comment https://github.com/Walstruzz/edge_eval_python/blob/7ca20a685d753f71bbcd6347b2b7bba54e22c32c/impl/edges_eval_dir.py#L37 rewrite as ``` python gt = [cv2.imread(g, cv2.IMREAD_UNCHANGED) for g in gt] ``` comment https://github.com/Walstruzz/edge_eval_python/blob/7ca20a685d753f71bbcd6347b2b7bba54e22c32c/impl/edges_eval_dir.py#L156 rewrite as ``` python...
#1 I will check it soon.
Thanks for your feedback, code updated.
感谢反馈，已更新。
根据阈值迭代的时候主要是算了这个 https://github.com/Walstruzz/edge_eval_python/blob/3e2a532ab939f71794d4cc3eb74cbf0797982b4c/impl/correspond_pixels.py#L427
这个倒不太清楚，它在eval的时候其实是当某个点在一个小邻域存在gt中的像素点时则认为预测对了，而不是直接算&那么严格。
see [protobuf issue 5868](https://github.com/protocolbuffers/protobuf/issues/5868) * **open Makefile.config** * **add protobuf include path in `INCLUDE_DIRS`** * **add protobuf lib path in `LIBRARY_DIRS`** It works for me. :)
> @Walstruzz Walstruzz > Could you please tell me the Makefile.config file setting details? I have installed the protobuf in my system. > And the `protoc --version` command is worked....
@Yachuan-Li 感谢关注，这部分代码是参考的 https://github.com/davidstutz/extended-berkeley-segmentation-benchmark 进行实现。 https://github.com/Walstruzz/edge_eval_python/blob/add539b95e9222ba9426924decee1c61ebcefb83/impl/correspond_pixels.py#L1-L19
I re-implement it. https://github.com/Walstruzz/edge_eval_python