GA-SVM Optimization Algorithm for Tracking Moving Objects

Moving target tracking combines the previous state estimation information with the real-time measurement information provided by the sensor, and uses signal and information processing methods to track the target. Due to its applications in robot visual navigation, public safety monitoring, intelligent traffic monitoring, military navigation and air defense missiles, it has attracted great attention in the world, and promoted the development and improvement of moving target tracking algorithms. This paper studies GA-SVM (Genetic algorithm-support Vector Machines) optimization algorithm, state estimation and fusion, hybrid system modeling and estimation, random matrix modeling and estimation and other theories and methods. Most of the early target tracking methods are based on models. With the rise of various algorithms, the establishment of state-space-based scale models has become an important research direction. The emergence of new aircraft and new sensors has also injected new connotation into the difficult problem of target tracking, and promoted the development and improvement of the theory. The experimental results show that the value of MOTA (multiple object tracking accuracy) for motion object tracking based on GA-SVM algorithm is $86.4 \%$.