论文部分内容阅读
为了保证智能车辆在交叉口内的诱导行驶,提出了一种交叉口信号灯检测与识别方法.首先利用Hough变换检测交叉口内的停车线,然后采用颜色空间变换检测红、黄、绿三色信号灯,最后建立自联想存储器以识别切分出来的信号灯时间字符.通过20个实际交叉口场景测试数据验证,采用所提出的方法,停车线检测正确率达90%,信号灯检测正确率为85%,在信号灯字符正确分割出来的基础上,字符的识别率达97%.结果表明提出的方法能够十分有效地进行数字信号灯的检测,并具有足够的鲁棒性识别“破损”及带噪声的字符.
In order to ensure the intelligent vehicles in the intersection of induced driving, a method of signal detection and recognition of intersections is proposed.Firstly, the Hough transform is used to detect the parking lines in the intersections, and then the color space is used to detect the red, yellow and green signals, finally Established self-associative memory to identify the segmented semaphore time characters.According to the 20 actual intersection scene test data validation, using the proposed method, the detection accuracy of parking line is 90%, the correctness of signal detection is 85% Based on the correct character segmentation, the recognition rate of characters is 97% .The results show that the proposed method can detect digital lights very effectively and has enough robustness to recognize “broken ” and noisy characters.