Owen-Liuyuxuan / visualDet3D

Official Repo for Ground-aware Monocular 3D Object Detection for Autonomous Driving / YOLOStereo3D: A Step Back to 2D for Efficient Stereo 3D Detection
https://owen-liuyuxuan.github.io/papers_reading_sharing.github.io/3dDetection/GroundAwareConvultion/
Apache License 2.0
365 stars 77 forks source link

请问大佬使用的双目匹配方法和stereo3d.md文件的作用 #78

Closed Chain-Mao closed 1 year ago

Chain-Mao commented 1 year ago

您好,麻烦问下大佬,项目中的双目匹配用的什么算法,我看代码中用到了opencv的BM块匹配算法?还有麻烦问下代码中stereo3d.md文件及其生成的imbd.pkl文件起到什么作用?

Chain-Mao commented 1 year ago

Opencv的BM算法的精度并不是很高,作者大佬在匹配后做了优化吗

Owen-Liuyuxuan commented 1 year ago

docs/stereo3d.md is the document I wrote to demonstrate the pipeline of using the code on stereo 3d detection training and testing.

imdb.pkl are read in and used in dataset class.

I didn't further optimized the BM algorithm. The result of the BM algorithm is used to help the training of stereo features but not directly affecting the result of 3D prediction.

Chain-Mao commented 1 year ago

docs/stereo3d.md is the document I wrote to demonstrate the pipeline of using the code on stereo 3d detection training and testing.

imdb.pkl are read in and used in dataset class.

I didn't further optimized the BM algorithm. The result of the BM algorithm is used to help the training of stereo features but not directly affecting the result of 3D prediction.

docs/stereo3d.md is the document I wrote to demonstrate the pipeline of using the code on stereo 3d detection training and testing.

imdb.pkl are read in and used in dataset class.

I didn't further optimized the BM algorithm. The result of the BM algorithm is used to help the training of stereo features but not directly affecting the result of 3D prediction.

感谢作者回复,助您新年快乐!还有我想请问下,您在推理时使用双目匹配的方法了吗,如何在得到双目3D推理结果框的同时得到双目视差图呢?