| 摘要 |
<P>PROBLEM TO BE SOLVED: To provide an image processing system for improving the recognition system of a lane mark. <P>SOLUTION: In this image processing system, a lane mark candidate is recognized by a first processing means 110 based on the luminance of each pixel of a road surface image. Also, a first index expressing the continuity of the edge of the lane mark candidate and a second index expressing the matching of the width of the lane mark candidate and the specific width of the lane mark and a third index expressing the uniformity of the luminance of pixels included in the lane mark candidate. Then, the lane mark candidate which is highly likely to be pertinent to the true lane mark is recognized as a lane mark by a second processing means 120 based on the first, second and third indexes. <P>COPYRIGHT: (C)2007,JPO&INPIT |
