Research On Trajectory Acquisition And Activity Understanding Of Running Vehicles

With the rapid development of computer technology, more and more researchersstart to research the field of computer vision. In this field, object tracking and activityunderstanding has become a hot research topic. many researchers have done manytentative research, and obtained many meaningful research achievements. In thispaper, we mainly discussed the application and research of SURF algorithm in objecttracking. For the activity understanding, I also do some research under the conditionsof setting a specific background.For the object tracking, the SURF algorithm was used to research it in this paper,however, the integral image computing algorithm used by SURF has problems ofcosting more time and mismatching of image matching, expecially the computationtime and matching time. In order to solve this problem, integral image computingalgorithms and matching algorithms were improved in this paper. Firstly, featuredetection and pre-matching of images are done by using SURF algorithm. Secondly,the mismatching is wiped out by using scale restrict, Finally, the matching processwere completed effectively and accuratly. For the activity understanding, the researchwas based on a bustling Transportation crossroads in this paper, and some rules of thebehavior of the moving vehicle was developed, such as urn left, turn right, go straightand turn around. The experimental results demonstrate that this improved method Caneffectively and real-timely to complete the process of the object tracking and theactivity understanding of the moving vehicle．...