基于三角函数的加权平均全景环视图像融合方法

Weighted average around view monitor image fusion method based on trigonometric function

全景环视(AVM)系统是减小汽车盲区的有效工具,针对已有AVM图像中存在的拼接缝以及拼接误差等问题,提出了两种依据像素坐标的三角函数值计算权值的像素级图像融合方法。对于短距AVM图像的融合,提出了一种以融合区域像素坐标的正弦/余弦值为融合权重的图像融合方法,命名为“时钟扫描图像融合法”;对于长距AVM图像的融合,提出了一种依据正切函数曲线拟合矩形融合区域,并利用正切函数系数构造权值区间确定各像素权重系数的方法,以同时顾及3条融合边界,称之为“tan—加权图像融合法”。研究结果表明,两种图像融合方法在消除相邻图像之间的拼接缝、提升融合图像的清晰度等方面具有良好效果,可实现融合区域的平滑过渡。

The around view monitor(AVM)system is an effective tool to reduce blind spots of automobiles.Aiming at the problems of splicing seams and splicing errors exist in AVM images,two pixel-level image fusion methods are proposed,which calculate weights according to trigonometric function values of pixel coordinates.For the fusion of short-range AVM images,an image fusion method based on sine/cosine values of the pixel coordinates of the fusion region as the fusion weight is proposed,which is named"clock scan image fusion method".For the fusion of long-range AVM images,a method is proposed to fit the rectangular fusion region according to the tangent function curve,and using the tangent function curve to construct a weight interval to determine the weight coefficient of each pixel,so as to take into account the three fusion boundaries at the same time,which is called"tan-weighted image fusion method".The research results show that the two image fusion methods have good effects in eliminating the splicing seams between adjacent images and improving the clarity of the fused images,and can achieve a smooth transition of the fusion regions.