In recent years, polarization correlated imaging has been widely studied by researchers. It has great practical value in the field of target detection and feature extraction. A full Stokes polarization correlated imaging system can obtain multiple polarized images of the target. Using these images, the intrinsic polarization information of the target can be analyzed from different views. However, there is a strong complementarity and redundancy between these images. Therefore, HSL-RGB image fusion technology was applied to this system. The number of polarized images acquired by the system was effectively merged to comprehensively describe the target structure and improve the target detection and recognition efficiency. The experimental results confirm that the fusion technique is effective in improving the recognition and detection performance of polarization correlated imaging system.