Compared to the discrete image recording method during the assembly process of manned spacecraft, panoramic imaging technology has the advantages of high resolution and wide viewing angle. However, the application of traditional panoramic image acquisition and processing methods in the image recording during the assembly process of manned spacecraft can lead to issues such as low coverage of extravehicular panoramic images and insufficient image registration accuracy. According to the structural characteristics of manned spacecraft, this paper proposes a method for acquiring and processing extravehicular panoramic images of manned spacecraft, designs a layout scheme of extravehicular panoramic images, develops a control-by-wire panoramic pan-tilt-zoom(PTZ) system, and proposes an image registration algorithm based on perceptual hash matching point purification. The perceptual hash matching point purification method proposed in this paper calculates the similarity of small area images around two matching points based on hash fingerprints, and removes matching point pairs that are less than the set threshold. The method proposed in this paper meets the remote acquisition requirements of extravehicular manned spacecraft above 15 meters, ensuring complete coverage of extravehicular panoramic images, and improving the quality of panoramic images.