The Application Of Computer Vision In Excavator Robot Walking

The history and status of computer vision and intelligent excavator robot are introduced in this paper.In order to solve the problems met during autonomous driving and digging of an excavator robot,road identification algorithm and image matching algorithm are studied.This paper includes several parts as follows:1 Computer vision system and image processing are introduced and then the major parts are selected and located.2 A road identification algorithm is proposed.Blocks are used when the image is enhanced. None-road region is figured out by characteristic points.The road edges are decribed by linear fitting.3 A general image matching algorithm is proposed.It uses relative position of intensity to cover the disadvantages of classical image matching algorithms.It is based on the relative position of neighbor pixels combined with partial matching algorithm.The proposed algorithm is robust and can be carried out easily.4 The theory and mathematical model of binocular stereo vision based measurement are introduced in detail.By using the simplified parallel optical axis model,the feasibility and reliability of this algorithm are demonstrated.