OpenVehicleVision
OpenVehicleVision copied to clipboard
An opensource lib. for vehicle vision applications (written by MATLAB), lane marking detection, road segmentation
参数 | Dataset | param | | --- | --- | | roma | | | BDXD54 | 0.2 | | BDXN01 | | | LRAlargeur13032003 | 0.11 | |...
Image Mosaicing in MAtlab using SIFT and RANSAC VLFeat http://www.vlfeat.org/applications/sift-mosaic-code.html 任务: 每帧检测完成后,执行与上一帧的拼接,并记录参数,全部拼接完成后,生成三维路面图,并估计运动轨迹,重建每一帧图像。对比。
标注显示工具
- [ ] 多类别的GT - [ ] 单个图片处理和视频处理 - [x] 突出GT边缘 - [ ] 结果对比,和GT对比的漏报和错检 - [ ] 多个结果对比,多的漏报(越少越好)和多的错检(越少越好) 三种方式显示GT标注,目前仅支持两类识别的GT 
下一步工作
转向 python+opencv 读 kitti eval 评估的py代码 阅读 VLfeat 源码 TODO https://github.com/baidut/OpenVehicleVision/search?utf8=%E2%9C%93&q=TODO
# 2D Hist [灰度图转3d](http://www.mathworks.com/matlabcentral/fileexchange/29485-meshcanopy)  [柱状图](http://cn.mathworks.com/matlabcentral/fileexchange/46012-2d-weighted-histogram) [柱状图+散点](http://cn.mathworks.com/matlabcentral/fileexchange/45325-efficient-2d-histogram--no-toolboxes-needed)  用球显示3d 直方图[3D histogram of RGB image](http://www.mathworks.com/matlabcentral/fileexchange/38685-3d-histogram-of-rgb-image)  [用矩形显示3d 直方图,效果较次](http://cn.mathworks.com/matlabcentral/fileexchange/52686-representation-of-the-3d-joint-rgb-histogram) [散点RGB 图](http://cn.mathworks.com/matlabcentral/fileexchange/52687-rgb-color-representation)  RGB/HSV Distribution http://cn.mathworks.com/matlabcentral/fileexchange/49002-rgb-hsv-distribution
方案1 exe 调用 ``` Raw = imread('F:\Documents\pku-road-dataset\1\EMER0009\0379.jpg'); Seg = FelzenSegment(Raw); implot(Raw,Seg); ``` 方案2 mex 编写 方案3 c实现 matlab image segmentation kmeans [Fast kmeans Algorithm Code](http://cn.mathworks.com/matlabcentral/fileexchange/44598-fast-kmeans-algorithm-code)
路面提取算法评测
# Related Papers Road Detection Based on Illuminant Invariance dataset http://www.cvc.uab.es/adas/databases
十字路口,路面车辆干扰 Filtering and Smoothing Data http://cn.mathworks.com/help/curvefit/smoothing-data.html