| Recently,couple with the increase of the expressway's length and vehicle stock,the number of traffic accidents is rising , the road traffic problems have become serious in our nation by threatening people's life and assets. Collision Warning System (CWS),which is a part of the vehicle initiative safety system,is aiming at warning drivers when the car is in dangerous situations.The technology research increasingly attracts the attention of scholars worldwide.The paper chooses the visual sensor as the CWS sensor and makes a further study in the following four points based on various papers concerned home and abroad. .(1) The establishment of the critical safety distance model.Firstly,the braking process is analyzed, which considers the effect of the relative velocity. Secondly,the variation of the maximal deceleration in different road situations is studied .Finally, the critical safety distance model is built .The result from the simulation shows that the critical safety distance from the model is more reasonable.(2) The improvement of the vehicle detection algorithm.The symmetry characteristic of the horizontal edges and the feature of road vanishing points are used to determine regions of interest,which is based on the analysis of multi-cues fusing vehicle detection algorithm. In the regions of interest, the characteristic of vehicle shadow is used to determine the vehicle candidate regions. In the next step, the symmetry characteristic of the vertical edges is used to remove the false candidate regions. Experimental results show that the method is not only improving the real-time but also ensuring the accuracy of detecting.(3) The visual distance measuring method is presented. The relationship between the onboard coordinate and image coordinate is analyzed, and the translation function is got.According the camera's inner parameters from camera calibration and the extrinsic parameters from the real scene experiment,the real world's distance is computed.The error between the test data and the real data is analyzed.The result indicates that the error is inside the allowable area,which proves the method can satisfy the requirement of CWS.(4) The test of the system's validity .In order to prove the validity of the CWS ,the test is implemented on Nanjing-Hangzhou Highway, the result shows the system has good reliability. |
