Design Of Spatial Dynamic Target Measurement And Tracking Algorithm Based On Stereo Vision

Computer vision attempts comprehending surroundings through the analysis and understanding of images.The ability to perceive enables machines to interpret the threedimensional structure of their surroundings and allows them to navigate through the environment.In computer vision,depth perception is achieved by stereo matching,a process that identifies correspondences between pixels in image pairs acquired from horizontally offset cameras.It is possible to calculate depths from disparities between pixels in correspondence.A stereo matching method allows for recovery of highly accurate disparities in real-time.This method combines a pixel dissimilarity metric computed using both the gradients and the census transforms of the input images,a non-iterative local disparity selection scheme based on an efficient approximation of the edge-preserving bilateral image filter,and a refinement technique that iteratively improves the accuracy of disparities.The refinement technique,which also benefits from the use of the bilateral filter,eliminates mismatches by penalizing disparities that disagree with the the disparity estimates generated using local disparity values and gradients.When evaluated using the Middlebury stereo performance benchmark,the proposed method ranks 5th when using the training and test image sets,respectively,in terms of the overall accuracy of stereo matching measured as the average percentage of pixels with the absolute disparity error greater than 2 pixels at the nominal image resolution.Simultaneously,the method achieves the lowest error rates for 3 out of 15 image pairs in the training set,and 3 out of 15 image pairs in the test set.This method is also shown to enable robust matching in the presence of radiometric distortions caused by changes in illumination or camera exposure.The high accuracy of matching,that is largely maintained in the presence of radiometric distortions and the ability to operate in real time,makes the proposed method wellsuited for applications such as robotic navigation and structure reconstruction.A stereo vision system based on the proposed method to estimate the position,velocity and heading of the spatial dynamic target is designed.The system consists of stereo vision ranging,object detection and tracking,and trackingerror minimization.These errors are mainly due to image quantization limitations and pixel miscorrespondences in the stereo pixel-matching process.While more complicated matching algorithms may lead to a better depth reconstruction of the scene,it costs a lot of time.This system combined a simple stereo matching algorithm,along with a predictive-corrective approach based on the implementation of an Extended Kalman Filter(EKF),using suitable choices of probabilistic models representing the motion of the spatial dynamic target and the stereo measurements,in order to minimize the tracking errors related to the stereo measurements,and therefore,improve the accuracy of the state estimation of the target.Results from simulations reveal the effectiveness of the system to compute accurate estimations regarding the state of the target over a variety of motion conditions.