基于自适应阈值的车辆边缘提取算法研究

Research on Vehicle Edge Distilling Algorithm Based on Self-Adopted Threshold

实际道路视频图像背景复杂多变,给车辆边缘检测带来困难。提出用最大类间方差法求边缘检测算子的阈值,并给出具体迭代求解过程,得到的阈值能随图像背景的变化而变化,具有自适应性。实验结果表明了算法的可行性和有效性。

Vehicle edge-detection is quite difficult under the conditions of the real complex and volatile traffic image. A max interclass-deviation algorithm and the steps of solving process are proposed to get the value of edge-detection threshold. In this algorithm, the threshold is automatically adjusted by the images. Experimental results prove the feasibility and validity of the improved algorithm.