童车变形量图象处理中的霍夫变换的优化

Hough Transform Optimization in the Process of Image Disposal for Children's Vehicle Distortion

为了解决童车变形量图形处理过程中的计算效率问题,对霍夫变换进行优化。首先将连续搜索法融入到霍夫变换中,滤去了过大或过小的轮廓,对所获得的连续轮廓进行筛选,从而减少了霍夫图像变换点的数量,然后最大限度使用连续变换中的信息,变全图搜索为局部元素搜索,大大地缩小了搜索范围。实验结果表明:对于尺寸为640×180的图像进行霍夫变换,融入了连续搜索的霍夫变换的计算速度是原始的霍夫变换的13倍,融入了连续搜索的优化的霍夫变换的计算速度是原始霍夫变换的497倍,取得了较好的效果。

In order to improve the calculation efficiency in the process of image disposal for children＇s vehicle distortion,Hough transform was optimized.First,the continuous search was melted with the Hough transform,and the oversize less size outline of the image was filtrated,at same time the continuous outline obtained was screened.By this way,the points participating in the Hough transform might be decreased.Second the information in the continuous transform was maximized in application to change the whole diagram searches into the partial element search,which reduced the search range consumedly.Experiment shows：compared with the Primitive Hough transform,the calculation efficiency of the Hough transform melted with the continuous search is 13times of the original Hough and the calculating speed of optimized Hough transform melted with continuous search is 497times of the original.Thus a good result is obtained.