An Auxiliary Heterogeneous Visual Localization Algorithm Based on Improved Hough Transform Line Matching

With the increasing demand for all-weather, uninterrupted aviation missions, heterogeneous visual navigation and positioning technology based on visible images and infrared images has been developed, and the key technology of heterogeneous image matching has become a research hotspot. Due to the different imaging mechanisms of heterogeneous images, it is difficult to directly apply the homogeneous image matching algorithm to heterogeneous images, and the feature extraction accuracy of algorithms based on line segment structure such as Hough transform needs to be improved. Therefore, an improved Hough transform line matching algorithm to assist heterogeneous visual positioning is proposed in this paper. Based on the heterogeneous image obtained by different optical sensors, the adaptive Canny operator of the improved maximum inter-class variance method is used for edge detection. The detected edge binary image is processed by Hough transform, and the line features in the heterogeneous images are extracted for matching. The experimental results show that the Hough transform line matching algorithm using the improved Canny operator has a line feature detection accuracy of 66%, which is 74.23% higher than the traditional Hough transform line matching algorithm. The number of correctly matched line features of the proposed algorithm is better than that of the traditional algorithm, which can provide ideas for the pose calculation based on heterogeneous images. The running time of the proposed algorithm is better than that of the traditional Hough transform line matching algorithm, and the total time is reduced by 46.62%.