The global vision of robot soccer system is introduced firstly in this paper,then some affecting identify result factors are especially analyzed from hardware and software of the vision system,according to these factors ,some improvement methods was given below:Firstly, image distortion was solved by camera calibration based on BP Network;secondly,HSI color model was applied to implement the partition of object images; and thirdly ,a more reasonable and anti-jamming design was put forward by comparing and analyzing various representative color tag designs with a fast and relatively exact algorithm given correspondingly.Experiments show that the stability and accuracy of vision system are improved, the preparation time for competition is also reduced and achieved favorable effect in play . It has a bright future in application prospects. |