| 摘要 |
<P>PROBLEM TO BE SOLVED: To provide an image processor which can perform vehicle recognition more accurately from a photographed image taken on a road. <P>SOLUTION: An image processing section encodes perpendicularly/horizontally features of a feature extraction image created in a feature image creation section 51 in a perpendicular/horizontal feature coding section 54 based on photographed images of vehicles. As to this feature extraction data, the narrow width data or the like is combined by a horizontal feature combining section 61 based on a vehicle determination value of a vehicle width condition, and the wide width data is separated and corrected by a wide width data separation section 62. Then, vehicle tail sensing, vehicle length calculation, and vehicle body decision are carried out by a feature correction data in a vehicle recognition section 66. A vehicle tail position is computed by performing light sensing by using a background difference image for light sensing created in a background image creation section 52. The vehicle tail position is determined by consolidating vehicle tail sensing results of vehicle body analysis and light analysis in a vehicle tail decision section 69. <P>COPYRIGHT: (C)2006,JPO&NCIPI |
