| With the improvement of the living standard,vehicle has become an indispensable part of our daily life,and people are paying more and more attention to the safety of driving.In order to solve the problem of visual blind area which caused by many factors such as the structure of vehicle,Advanced Driver Assistant System has appeared.Advanced Driver Assistant System observe the road condition by using different sensors.The vehicle around-view monitor system is one of ADAS.The vehicle around-view monitor system obtains the surrounding image message through the four fish-eye cameras which install at the front、back、left and right side of the vehicle,then use ECU to process the algorithms,finally,display the bird-eye view of the mosaic image to the drivers.Aiming at the vehicle around-view monitor system,this paper focuses on the technologies and algorithms of the whole system,realizes the software and hardware architecture of the vehicle around-view monitor system,the main contents of this paper as follows:(1)Based on the research of fisheye camera imaging principle and fish eye image correction algorithm,a spherical projection algorithm based on scan line is proposed,and interpolation algorithm is applied to optimize the result.(2)By studying the principle of camera imaging and the algorithm of inverse perspective transformation,an inverse perspective projection algorithm based on the mark points is proposed.(3)Through the research on the image mosaic algorithm,we shear the four images before the image mosaic algorithm,then fuse the sheared images.Based on the traditional image mosaic algorithm,a block matching RANSAC algorithm is proposed to eliminate mismatching of images,finally use the weighted image fusion algorithm to eliminate the image seams.The experiment shows that the improved and optimized algorithm proposed in this paper can greatly simplify the operation steps and improve the efficiency of the algorithm,the whole experiment showed a good result. |
PDF Full Text Request