搜索资源列表
ball_catcher
- This includes the project using a stereo vision to catch the ball shot from a high pressure air cannon. The trajectory of the ball is located first by finding the ball color in the left and right camera. Then the trajectory is calculated to estimate
ZhouCVPR2010
- CVPR2010的一篇很牛的文章，通过在相机前段加一个滤镜，可以达到单幅图毫米级的深度估计-CVPR2010 very cattle in an article, by adding a filter preceding the camera can achieve a Single Figure millimeter depth estimation
Matlab
- 为了提高四足机器人在复杂环境下的适应性,重点研究了采用飞行时间(TOF)原理相机的四足机器人环境感知策略并改进了地形识别及路径规划算法.首先采用高斯过程回归(GPR)模型对TOF相机的距离数据进行误差校正,解决了采用传统多项式或三角函数模型进行误差修正时模型阶次过高及函数组合复杂的问题.基于得到的环境深度信息,采用数字高程模型(DEM)进行地形描述,并通过计算各栅格的坡度、粗糙度、起伏度对地形进行识别.粗糙度由该栅格所处的坡度平面与其8邻域高程点的离散程度进行计算,避免了采用高程方差计算时对粗糙
zhuan8
- 主要用于路径规划，程序读入之初，需要先读入Kinect相机拍摄的深度图像。(Mainly used for path planning, the program read at the beginning, you need to read the depth image taken by the Kinect camera.)
dijkstra1
- 用于路径规划，程序初始需要读入Kinect相机拍摄的深度图像(For path planning, the program initially needs to read the depth image taken by the Kinect camera)
DeconvolutionCode-LevinEtAl07_original
- 2007 ACM Image and Depth from a Conventional Camera with a Coded Aperture(Anat Levin)