研究目的
Addressing the problem of stereo vision-based environment mapping and optimal path planning for an autonomous mobile robot.
研究成果
The paper successfully re-examined and extended the methodology for stereo vision-based mapping and path planning, demonstrating effective obstacle avoidance and navigation in indoor environments using the A-Star algorithm with binary heap. Future work could involve testing in more dynamic environments and further optimization.
研究不足
The methodology is tested in indoor environments and may have limitations in dynamic or outdoor settings; computational efficiency depends on data trimming and algorithm parameters.
1：Experimental Design and Method Selection:
The methodology involves capturing stereo images, calibrating the camera, converting to edge images, computing disparity, generating a 3D point cloud, trimming the data (floor, ceiling, depth), building a 2D navigation map with occupancy grid cells, scoring cells based on point density and weight, and using A-Star with binary heap algorithms for path planning.
2：Sample Selection and Data Sources:
Real data acquired by the Innopolis autonomous mobile robot in indoor environments, using stereo images from a ZED camera.
3：List of Experimental Equipment and Materials:
ZED stereo camera, Innopolis mobile robot based on Traxxas 7407 RC Car Model, NVidia Jetson TX1 embedded controller, Lynx-motion quadrature motor encoder, MATLAB and C++ software.
4：Experimental Procedures and Operational Workflow:
Steps include image capture, calibration, disparity computation, point cloud generation, data trimming, map building, cell scoring, and path planning with A-Star algorithm.
5：Data Analysis Methods:
Comparison of block matching and semi-global block matching methods for disparity map generation, evaluation based on block size and frame per second, and analysis of A-Star algorithm performance.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
