一种运动车辆的阴影消除新方法

针对运动车辆阴影消除时可靠性和实时性难以兼容的问题,提出一种基于彩色检测线的线间差分的阴影消除新方法.通过检测运动区域四周方向线的线特性确定出运动区域所属的车辆/阴影模型,结合车辆/阴影模型和运动区域找到检测线,再利用检测线在彩色空间内的线间差分和检测线的属性找到阴影和车辆的分界点,进而根据分界点找到车辆和阴影的分界,从而实现阴影的消除.由于利用了检测线在彩色空间中的颜色特性和检测线在灰度空间中的背景图像,因而新方法具有较强的噪声抑制能力,并提高了低亮度车辆的阴影消除效果.实验结果表明,所采用的阴影消除新方法不仅具有较高的可靠性,也具有较好的实时性.

New Method of Shadow Elimination for Moving Vehicle

Focusing on the problem that the reliability and real-time requirement for eliminating cast shadow of moving vehicles are hard to be compatible, a new method based on the difference between trap lines in color space was proposed. Through detecting the line character of the four lines around the motion region, the vehicle/shadow model belonging to the motion region was determined. Combining vehicle/shadow model with motion region the trap line was found, and then the boundary points between vehicle and shadow were located by using the attribute of trap lines and the difference between trap lines in the color space. Furthermore, the shadow of the moving vehicle was eliminated by the boundary drawn from the boundary points. Because of using the color character and the gray background of the trap lines, it has stronger ability to restrain noise and improves the effect of shadow elimination for vehicles with low pixel intensities. The results verify that the new method is reliable, and its computation cost satisfies the real-time demands.