Abstract: The known current intelligent vehicle detection algorithms have problems such as large errors and inaccurate measurement results in practical applications． The algorithm used in this paper is based on the vehicle detection mechanism of the car hull line in the moving vehicle，and uses the feature of the bottom shadow to exist．And the binocular stereo matching distance measurement method based on machine vision is selected to realize the nearest vehicle distance measurement． Provide reliable information for intelligent vehicles and driver assistance．The final result of the system confirms the availability of more accurate vehicle and distance information in front of the road． The effective detection distance is 10 m ～50 m，and the overall recognition rate can reach 85% or more．

Key words: vehicle bottom dash, intelligent vehicle adaptive threshold segmentation, vehicle detection, binocular distance measurement