The Research Of Video Navigation Control Technology Based On The Freescale Intelligence Car

In all intelligent car key technology, Video navigation has the advantages of largeamount of information and flexible guiding, thus becoming the main research directionof the intelligent car. This thesis carries out the research based on the structured roadVideo navigation algorithm.Firstly, a deeply research of the four main intelligent car Video navigationalgorithms was taken. Through analysis and comparison, in the aspect of bend markline detection, there is a lack of strong robustness, high real-time, accuracy videonavigation algorithm. The thesis then improves the algorithm of region growing, andapplies an algorithm based on image restoration.Secondly, details of video navigation algorithm based on image restoration areexpounded. The collected images are binarized, the expansion, corrosion and noise aretreated, and background segmentation is implemented, thus obtaining a clearer lane.Then regional growth and image restoration are combined, to design a lane markingextraction algorithm based on the image restoration. In the description of the lane markline segment, the thesis applies the suitable fitting method based on judging the lane,and solves the equation of the lane.At last, to estimate the effectiveness of the algorithm, test it in Freescalemicrochip system and indoor lane situation.The result shows that video navigation algorithm based on image restoration hasthe advantage of accuracy, real-time and robustness in bend testing.