A Kind Of High-Accuracy Measure Algorithm For Motion Object

Velocity is one of the elementary parameters of object's movemenet. Nowadays as the rapid development of computer vision and image processing technique,the new methods have been gained to acquire the instantaneously motion parameters of moving object. In this research we put forward a kind of imprvoed optical flow method for dynamic measuring velocity.The motion status is sampled by high-spced area CCD camera and then the image information is transferred to computer, at last the velocity of moved object is analyzed with the method of image processing. Images sequences are been pretreatment by partition and filtering.According to the theory of motion analysis, we put forward a kind of improved optical flow algorithm based on the difference image method and the optical flow method. In this algorithm, a cursory move target is been gained by difference image method.Then the velocity is been calculated by cursory target through optical flow method.We throw away the background which caused the abnormal optical flow estimation after analyzed the optical flow field. We can not only reduce the calculation quantity, but also raised the credibility of the measurement.So it can also satisfied the requirement of the fast and high accuracy.During this research, we build up the experimental system of image measuring which simulate the condition of the practical field in the lab. We complete the software by use C language under the Visual C++ edit environment. We validate the improved optical flow algorithm by a series of experiments on this system...