Aiming at the problem that the registration process of infrared and visible images is greatly affected by the difference of gray level and the registration of feature points is difficult,an image registration algorithm based on saliency detection and ORB feature points is proposed. Firstly,the optimized HC-GHS saliency detection algorithm is used to obtain the saliency map of the image. Secondly,ORB algorithm was used to detect the feature points on the saliency map,and the robust feature points were screened by Taylor series,and the grouping matching strategy was carried out according to the direction of feature points. Finally,hamming distance is used to match the feature points. Experiments show that the proposed algorithm can achieve the registration between the infrared and visible images accurately,and it has good effect under the infrared noise interference and scale changes.