Description: this document explains the process of training a pedestrian detector using Viola and Jones’
cascaded classifier approach. it was originally Platform: |
Size: 132096 |
Author:lordthanatos |
Hits:
Description: opencv训练的行人目标图像，有兴趣可以-opencv pedestrian training target image, are interested can look at Platform: |
Size: 53723136 |
Author:yuuu |
Hits:
Description: 这是一个行人样本的裁剪程序，能批量裁剪图像用于训练行人样本的训练，有个说介绍了使用方法，只是个测试版本，有点漏洞的。-This is a sample of the cutting process pedestrians can crop the image used for training the pedestrian volume of training samples, there is a method that describes how to use, just a test version, a little vulnerability. Platform: |
Size: 10001408 |
Author:shanleo |
Hits:
Description: 该代码用于各种图像的HOG特征提取，以进一步训练分类器进行行人检测，不错的-HOG for a variety of image feature extraction, in order to further training classifier Pedestrian Detection Platform: |
Size: 4096 |
Author:lili |
Hits:
Description: 用matla实现的行人检测，使用hog+adaboost的方法，内附程序运行时所需的大量训练及检测图片-Pedestrian Detection with matla achieved, the use of hog+ adaboost the method, enclosing the program is running a lot of training and testing images... Platform: |
Size: 51777536 |
Author:王龙 |
Hits:
Description: 提出一种多尺度方向(multi-scale orientation,简称 MSO)特征描述子用于静态图片中的人体目标检
测.MSO 特征由随机采样的图像方块组成,包含了粗特征集合与精特征集合.其中,粗特征是图像块的方向,而精特征
由 Gabor 小波幅值响应竞争获得.对于两种特征,分别采用贪心算法进行选择,并使用级联 Adaboost 算法及 SVM 训
练检测模型.基于粗特征的 Adaboost 分类器能够保证高的检测速度,而基于精特征的 SVM 分类器则保证了检测精
度.另外,通过 MSO 特征块的平移,使得所提算法能够检测多视角的人体.通过对于 MSO 特征块的装配,使得算法能
够检测人群中相互遮挡的人体目标.在INRIA公共测试集合及SDL多视角测试集合上的实验结果表明,算法具有对视角与遮挡的鲁棒性和较高的检测速度. -The multi-scale orientation (MSO) features for pedestrian detection in still images are put forwarded in
this paper. Extracted on randomly sampled square image blocks (units), MSO features are made up of coarse and
fine features, which are calculated with a unit gradient and the Gabor wavelet magnitudes respectively. Greedy
methods are employed respectively to select the features. Furthermore, the selected features are inputted into a
cascade classifier with Adaboost and SVM for classification. In addition, the spatial location of MSO units can be
shifted, are used to the handle multi-view problem and assembled therefore, the occluded features are completed
with average features of training positives, given an occlusion model, which enable the proposed approach to work
in crowd scenes. Experimental results on INRIA testset and SDL multi-view testset report the state-of-arts results on
INRIA include it is 12.4 times the faster than SVM+HOG method. Platform: |
Size: 1868800 |
Author:尹世荣 |
Hits:
Description: 视频图像的人群密度检测，多种人群密度场景下人群计数算法：
算法功能:建立图像特征和图像人数的数学关系
算法输入:训练样本图像1,2…K
算法输出:模型估计参数 ，参考图像
算法流程：1)对训练样本图像进行分块处理(算法1.1)；
2)通过算法1.2，计算训练样本各个对应分块的ALBP特征归一化，再用K-means算法(可使用opencv等算法库实现，不再描述其算法)，将图像块分成k(k<K)类，获取k(k<K)个聚类中心，即为参考图像；
3)对分块的图像进行与参考图像进行匹配。使用算法1.2求取ALBP特征，并求取其相似度 ，将相似性集合作为新特征并形成一个归一化的新特征 。
4)按照行人面积占图像块面积的比例，以60 为分界，分布采用径向基核函数 和线性核函数 。K(xi,x)建立图像特征和图像人数的SVR(支持向量回归机)模型可使用opencv中的SVM或libsvm，输出模型估计参数 。
-Population density detection of video images, a variety of crowd density scenes crowd counting algorithm:
Algorithm functions: a mathematical relation between the image features and the number of images
Algorithm Input: training sample image 1,2 ... K
Algorithm output: model estimation parameters, reference image
Algorithmic process: 1) the training sample image into blocks (algorithm 1.1)
2) by 1.2 algorithm to calculate the corresponding training samples of each block ALBP features normalized, then K-means algorithm (algorithm can be used opencv library implementation, no longer describe the algorithm), the image block is divided into k (k <K) class, gets k (k <K) clustering centers, namely the reference image
3) conduct of image block matching with the reference image. 1.2 ALBP characterized using an algorithm to strike, and strike the similarity, the similarity of a set of new features and forming a normalized new features.
4) pedestrian area accounted for in Platform: |
Size: 4759552 |
Author:徐云华 |
Hits:
Description: 这个是一个SVM分类器，可以再行人检测时用来分类训练样本，再MATLAB中直接调用-This is an SVM classifier, when pedestrian detection can be used to classify the training sample, and then directly call MATLAB Platform: |
Size: 1332224 |
Author:Jeans Load |
Hits:
Description: hog训练和分类的程序,hog特征可以用于行人检测，这个程序包括训练和分类
欢迎下载-hog training and classification procedures, hog feature can be used for pedestrian detection, this program includes training and classification
Welcome to download Platform: |
Size: 2048 |
Author:yql |
Hits:
Description: 使用HOG特征和SVM分类器来进行行人检测，通过svm训练分类器得到人的特定参数-Pedestrian detection using HOG features and SVM classifier, and the specific parameters are obtained by SVM training classifier. Platform: |
Size: 301056 |
Author:王慧 |
Hits:
Description: 基于HOG+SVM行人检测，有训练好的文件，也有训练的过程，精读的话一般般，提供一种检测的思路-HOG+ SVM-based pedestrian detection, trained good document, but also the training process, intensive reading, then a general, to provide a detection idea Platform: |
Size: 24918016 |
Author:李杰 |
Hits:
Description: 行人检测HOG+SVM进行图片中行人检测，提供训练用的pos和neg样本，效果还可以；没有SVM工具箱的，压缩包里已经提供了，安装一下即可(Pedestrian detection HOG + SVM for pedestrian detection in pictures, providing POS and neg samples for training, the effect is good; without SVM toolbox, the compression package has been provided, just install it.) Platform: |
Size: 9630720 |
Author:子夜o星空 |
Hits:
Description: 人搜索
依赖
在python3.7 MacOS 10.14.6下测试
——python包
——opencv-python
——tb-nightly
-torch>= 1.0
下载weights
从这里下载，密码:qscx
下载后，将权重放到文件夹 person_search_demo/weights中
测试
cd <指向此floder>的路径
python search
结果将保存在输出文件夹中
训练
您可以直接使用原始的YOLO代码进行培训。
行人再识别模型采用强里德基线模型。(Person search
Dependencies
Tested under python3.7 MacOS 10.14.6
- python packages
- opencv-python
- tb-nightly
- torch >= 1.0
Download weights
Download from here, password:qscx
After download, put the weights into the floder person_search_demo/weights
Test
cd <path to this floder>
python search.py
The results will be saved in the output folder
Train
You can directly use the original YOLO code for training.
Pedestrian re-identification model is adopted strong Reid baseline model.) Platform: |
Size: 5827584 |
Author:Kyfafyd |
Hits:
Search - pedestrian training