Extended Kalman-based multi-sensor fusion

With the continuous development of sensor technology and electronic information technology, data fusion technology is of great significance in improving the system's ability of detecting, recognizing and tracking targets. That is, a single sensor can only provide partially imprecise information, thus not excluding ambiguity in the description of the external environment. Given the complementarity exhibited by the differences in the performance of the sensors, synthesizing the information from individual sensors results in more accurate and reliable information than a single sensor can provide. A multi-sensor fusion method based on extended Kalman is proposed to address this problem. Experimental results show that the accuracy and robustness of target detection and tracking can be significantly improved by the fusion of infrared and millimeter wave radar data.