{"id":13436738,"url":"https://github.com/shaoshengsong/DeepSORT","last_synced_at":"2025-03-18T21:31:01.117Z","repository":{"id":37432130,"uuid":"185739303","full_name":"shaoshengsong/DeepSORT","owner":"shaoshengsong","description":"support deepsort and bytetrack MOT(Multi-object tracking) using yolov5 with C++ ","archived":false,"fork":false,"pushed_at":"2023-07-14T16:48:01.000Z","size":81,"stargazers_count":824,"open_issues_count":46,"forks_count":192,"subscribers_count":7,"default_branch":"master","last_synced_at":"2024-10-27T20:21:09.278Z","etag":null,"topics":["deepsort","track"],"latest_commit_sha":null,"homepage":"","language":"C++","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/shaoshengsong.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,"governance":null}},"created_at":"2019-05-09T06:26:54.000Z","updated_at":"2024-10-27T14:08:48.000Z","dependencies_parsed_at":"2023-10-21T14:00:25.444Z","dependency_job_id":null,"html_url":"https://github.com/shaoshengsong/DeepSORT","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/shaoshengsong%2FDeepSORT","tags_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/shaoshengsong%2FDeepSORT/tags","releases_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/shaoshengsong%2FDeepSORT/releases","manifests_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/repositories/shaoshengsong%2FDeepSORT/manifests","owner_url":"https://repos.ecosyste.ms/api/v1/hosts/GitHub/owners/shaoshengsong","download_url":"https://codeload.github.com/shaoshengsong/DeepSORT/tar.gz/refs/heads/master","host":{"name":"GitHub","url":"https://github.com","kind":"github","repositories_count":244310498,"owners_count":20432549,"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":["deepsort","track"],"created_at":"2024-07-31T03:00:51.745Z","updated_at":"2025-03-18T21:30:58.184Z","avatar_url":"https://github.com/shaoshengsong.png","language":"C++","funding_links":[],"categories":["C++"],"sub_categories":[],"readme":"# DeepSORT\n\n# MOT(Multi-object tracking) using yolov5 with C++ support deepsort and bytetrack\n\n\nflyfish\n\n## 前言\n代码采用C++实现，目标检测支持YOLOv5 6.2,跟踪支持deepsort and bytetrack。\n检测模型可以直接从YOLOv5官网，导出onnx使用\n特征提取可以自己训练，导出onnx使用，onnxruntime cpu 推理，方便使用.\n特征支持自定义维度例如 128,256,512等\n\n本文源码地址\n\n```c\nhttps://github.com/shaoshengsong/DeepSORT\n```\n\n## deepsort v1.12\n新增bytetrack跟踪\n\nbytetrack论文\n```c\nhttp://arxiv.org/abs/2110.06864\n```\n\nbytetrack代码\n```c\nhttps://github.com/ifzhang/ByteTrack\n```\n\n## deepsort v1.1\ndeepsort原论文地址 \n\n```c\nhttps://arxiv.org/pdf/1703.07402.pdf\n```\n\n\n```c\nMOT using deepsort yolo5 with C++\n```\n\n操作系统：Ubuntu 18.04\n### 版本更新说明\n\n去除了TensorFlow依赖\n为了不依赖硬件GPU，无需cuda，cudnn，更容易编译，使用PC版本。\n为了更方便编译，采用CMakeList.txt。\n\n\n### 依赖的库\nopencv，可以下载opencv-4.6编译安装\nEigen3安装\n\n```c\nsudo apt-get install libeigen3-dev\n```\n\nonnxruntime，可以直接解压使用，无需编译\n目标检测模型下载地址\n\n```c\nhttps://github.com/ultralytics/yolov5\n```\n\n网盘中有已经导出完成的模型\n\n### 文件下载\n百度网盘 \n链接：`https://pan.baidu.com/s/1igjNK2ty-H5AU_Ut08pkoA` \n提取码：0000\n内容包括\n\n```c\ncmake-3.21.4-linux-x86_64.tar.gz  \nonnxruntime-linux-x64-1.12.1.tgz\ncoco_80_labels_list.txt           \nopencv-4.6.0.zip\nDeepSORT                          \nyolov5s.onnx\nfeature.onnx                      \nyolov5x.onnx\n```\n\n\n### 使用方法\n#### 1 onnxruntime\n设置自己的onnxruntime的解压目录\n\n```\nset(ONNXRUNTIME_DIR \"/home/a/lib/onnxruntime-linux-x64-1.12.1\")\n```\n\n\n#### 2 模型配置\n以下三项根据自己的需要更改\n文件`tracker/deepsort/include/dataType.h`\n```c\nconst int k_feature_dim=512;//feature dim\nconst std::string  k_feature_model_path =\"./feature.onnx\";\nconst std::string  k_detect_model_path =\"./yolov5s.onnx\";\n```\n\n#### 3 主函数\n选择打开视频文件或者视频流等\n\n```c\ncv::VideoCapture capture(\"./1.mp4\");\n```\n\n### 扩展方式\n1 整体分为两部分，新增检测模块放置detector文件夹，新增跟踪模块放置tracker文件夹\n\n## deepsort v1.0\n### MOT using deepsort yolo3 with C++\n操作系统：Ubuntu 18.04\n编译环境：Qt 5.12.2\n深度学习的模型分两块，一个是目标检测，另一个是目标跟踪\n#### 目标检测的模型\n地址：`https://pjreddie.com/darknet/yolo/`\n\n\n#### 目标跟踪模型\nmars-small128 \nOpenCV DNN加载YOLO模型，不依赖Darknet库，cuda，cudnn\n依赖Tensorflow，目标跟踪的特征部分使用TensorFlow C++的api。\n\nOpenCV的安装可以参考\n\n\n地址:  `https://blog.csdn.net/flyfish1986/article/details/89157368`\n\n\nTensorflow的安装可以参考\n\n地址：`https://blog.csdn.net/flyfish1986/article/details/89406211`\n\n\n\n\n[多目标跟踪论文 Deep SORT 解读](https://flyfish.blog.csdn.net/article/details/89852370)  \n[多目标跟踪论文 Deep SORT 实现](https://flyfish.blog.csdn.net/article/details/90034289)  \n[多目标跟踪论文 Deep SORT 数据集说明](https://flyfish.blog.csdn.net/article/details/90070639) \n[多目标跟踪论文 Deep SORT 特征提取CNN Architecture](https://flyfish.blog.csdn.net/article/details/90642532)  \n[多目标跟踪论文 Deep SORT 特征训练PyTorch实现](https://flyfish.blog.csdn.net/article/details/90702620)              \n[多目标跟踪论文 Deep SORT 特征训练TensorFlow实现](https://flyfish.blog.csdn.net/article/details/90379444)  \n[多目标跟踪论文 Deep SORT 评测指标](https://flyfish.blog.csdn.net/article/details/90200171)  \n[匈牙利算法](https://flyfish.blog.csdn.net/article/details/104298521)  \n[卡尔曼滤波 - 方程组转换为矩阵形式](https://flyfish.blog.csdn.net/article/details/118635703)  \n[卡尔曼滤波 - 一个方程背后的样子](https://flyfish.blog.csdn.net/article/details/118636055)  \n[卡尔曼滤波 - 匀变速直线运动](https://flyfish.blog.csdn.net/article/details/118613382)  \n[卡尔曼滤波 - 冥冥之中自有定数的正态分布](https://flyfish.blog.csdn.net/article/details/116067569)  \n[卡尔曼滤波 - 数据融合 data fusion](https://flyfish.blog.csdn.net/article/details/118613307)  \n[卡尔曼滤波 - 当前均值与上一次均值的关系](https://flyfish.blog.csdn.net/article/details/117931292)  \n[卡尔曼滤波 - 状态空间模型](https://flyfish.blog.csdn.net/article/details/118636364)  \n[卡尔曼滤波 - 5个公式出现的顺序](https://flyfish.blog.csdn.net/article/details/118709808)  \n\n\n","project_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fshaoshengsong%2FDeepSORT","html_url":"https://awesome.ecosyste.ms/projects/github.com%2Fshaoshengsong%2FDeepSORT","lists_url":"https://awesome.ecosyste.ms/api/v1/projects/github.com%2Fshaoshengsong%2FDeepSORT/lists"}