Development of Algorithm for the Vision-based fast Object Recognition and Distance Calculation of the Soccer Robot
- Resource Type
- Dissertation/ Thesis
- Authors
- 요희 / Yao, Xi
- Source
- Subject
- object recogniton
soccer robot
- Language
- English
Soccer robot has become more and more popular in these years. To meet the necessity of fast vision process of the soccer robot, this paper presented an implementation of soccer robot using on stereo vision-based fast object recognition and distance calculation algorithm. Soccer robot vision algorithm consists of four steps. The first step is to do the color space transformation and classify each pixel in a color image into one of a discrete number of color classes and connect the regions of the same color and overlap under four-connectedness. The second step is to do the color segmentation and we could easily find the region of interest with the specified color, which is the object we are searching for and then we could get the contour and the centric of the object and do the object tracking. The third step is to do the stereo calibration of the two cameras the distance between the object and robot could be easily calculated using stereo vision. The fourth step is to design the hardware and to control the motion of the robot. We propose the color image segmentation method to find the object. According to the camera model and stereo vision, the distance between the object and the body of the robot could be calculated in real time using OpenCV.