中文摘要：

采用CCD摄像机采集高速公路场景图像，并通过图像颜色空间变换，将图像的RGB量值转换为色度-饱和度-亮度（HSV）量值。采用基于阈值的方法对场景图像中颜色饱和度分量进行二值化分割处理；利用场景二值化图像形状特征（周长、形状参数、圆形性参数）去除非目标区域，并通过搜索场景二值化图像方向投影值序列的突变点实现标志准确定位。采用HSV颜色模型中的亮度分量和最佳阈值法对场景图像中标志区域进行二值化处理。结果表明，应用上述方法能取得良好的效果。

英文摘要：

The CCD camera was used to shot the freeway scene images, and the HSV color space values were converted from RGB values through image color space transformation. Sub-variable S in freeway scene image binarization segmentation was processed based on the threshold value algorithm, and the shape features of the freeway scene binarization image, such as boundary length, form factor, circularity, were used to wiped off the non-target areas. By searching the catastrophe point of direction projection values sequence of freeway scene binarization image, the traffic sign region position was located correctly. Sub-variable V in HSV color model and the optimal threshold value algorithm were used to process sign area binarization in freeway scene image. Results show that this method is right.