{"id":17354390,"url":"https://github.com/meiqua/6dpose","last_synced_at":"2025-04-10T04:58:44.652Z","repository":{"id":44363248,"uuid":"125139301","full_name":"meiqua/6DPose","owner":"meiqua","description":"implement some algorithms of 6d pose estimation","archived":false,"fork":false,"pushed_at":"2019-08-22T13:07:46.000Z","size":74410,"stargazers_count":272,"open_issues_count":5,"forks_count":79,"subscribers_count":12,"default_branch":"master","last_synced_at":"2025-04-03T01:11:08.547Z","etag":null,"topics":["6d","estimation","icp","linemod","ork","pose","segmentation"],"latest_commit_sha":null,"homepage":null,"language":"Python","has_issues":true,"has_wiki":null,"has_pages":null,"mirror_url":null,"source_name":null,"license":null,"status":null,"scm":"git","pull_requests_enabled":true,"icon_url":"https://github.com/meiqua.png","metadata":{"files":{"readme":"README.md","changelog":null,"contributing":null,"funding":null,"license":null,"code_of_conduct":null,"threat_model":null,"audit":null,"citation":null,"codeowners":null,"security":null,"support":null}},"created_at":"2018-03-14T01:49:42.000Z","updated_at":"2025-02-26T02:18:05.000Z","dependencies_parsed_at":"2022-08-12T11:03:20.413Z","dependency_job_id":null,"html_url":"https://github.com/meiqua/6DPose","commit_stats":null,"previous_names":[],"tags_count":0,"template":false,"template_full_name":null,"repository_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/meiqua%2F6DPose","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/meiqua%2F6DPose/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/meiqua%2F6DPose/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/meiqua%2F6DPose/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/meiqua","download_url":"https://codeload.github.com/meiqua/6DPose/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":248161265,"owners_count":21057554,"icon_url":"https://github.com/github.png","version":null,"created_at":"2022-05-30T11:31:42.601Z","updated_at":"2022-07-04T15:15:14.044Z","host_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub","repositories_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories","repository_names_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repository_names","owners_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners"}},"keywords":["6d","estimation","icp","linemod","ork","pose","segmentation"],"created_at":"2024-10-15T17:19:58.034Z","updated_at":"2025-04-10T04:58:44.628Z","avatar_url":"https://github.com/meiqua.png","language":"Python","funding_links":[],"categories":[],"sub_categories":[],"readme":"\n# 6DPose\n\nimplement some algorithms of 6d pose estimation  \n\n- [x] linemod\n- [x] linemod levelup\n- [ ] linemod ros\n- [ ] latent class hough tree(LCHF)\n- [x] 3d convex segmentation  \n\nnote: LCHF is too hard to tune params so I'm not going to implement it...  \n[a better choice: patch_linemod](https://github.com/meiqua/patch_linemod)\n\n## prerequisite\n\n### pysixd\n\nfiles: params/  pysixd/  t_less_toolkit/  tools/  \ncopied from [sixd_toolkit](https://github.com/thodan/sixd_toolkit)  \ndeal with model reading/rendering, datasets reading and evaluation  \n\n### dataset\n\nget dataset under 6DPose folder using following cmd  \n```\nwget -r -np -nH --cut-dirs=1 -R index.html http://ptak.felk.cvut.cz/6DB/public/\n```\n\n### library\n\ninstall opencv3 with contrib rgbd module  \ninstall pybind11  \n\npip3 install -r requirements.txt\n\n### steps\n\nin target folder:  \nmkdir build  \ncd build/  \ncmake ..  \nmake  \n\nin 6dpose folder, if use pybind:  \npip3 install target_folder/  \n\npython3 target.py\n\n## linemod\n\nCodes in linemod.py will train and detect objects in downloaded dataset.  \nRefer to opencv linemod and ork linemod src  \n\n### result\n\n![image](./test/results/scene6_match.png)  \n![image2](./test/results/axis.png)\n\n## linemodLevelup\n\nLinemod suffers a lot from scale and occlusion problems, to deal wtih\nthem:  \n1. set low response closer to zero;(idea got from focal loss)  \n2. use depth histgram and 1D nms to find some primary scales  \n \nori linemod can't run while our features are more than 64, due to 8bit SSE.\nafter changing to 16bit SSE and preventing overflow,\nwe can have up to 8192 features now.\n\n\n[Chinese blog about the ideas](https://zhuanlan.zhihu.com/p/35683990)  \n\n[some tests](./linemodLevelup/readme.md)  \n\n## linemod ros  \n\nrun linemod with ros for real test, not done yet\n\n## latent class hough forest --- LCHF  \n\n- [x] linemod patch version\n- [x] hough forest\n- [ ] meanshift for leaf mode\n- [ ] icp for pose refine\n- [ ] train and test  \n\n[Chinese blog 0](https://zhuanlan.zhihu.com/p/35710562)  \n[Chinese blog 1](https://zhuanlan.zhihu.com/p/35740328)  \n\n## 3d convex segmentation\n\nsegment 3d cloud to convex part, core func have been done, try to deploy now  \n### result of cloud_seg.py\n![cloud](./cxx_3d_seg/test/cloud_seg.png)\n \nNOTE: To use pybind, static super4pcs must be built with flag -fPIC.\nBuild super4pcs using following lines:    \n```\nmkdir build\ncd build\ncmake .. -DCMAKE_CXX_FLAGS=\"-fpic\"\nmake\nsudo make install\n```\n[separated repo for deployment](https://github.com/meiqua/binPicking_3dseg)  \n\n[Chinese blog](https://zhuanlan.zhihu.com/p/36419676)\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmeiqua%2F6dpose","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fmeiqua%2F6dpose","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fmeiqua%2F6dpose/lists"}