Computational spectral imaging technology with high light utilization for space target detection

Abstract

Spectral imaging technology can obtain a three-dimensional data cube of the target, which has the advantage of "unification of maps". Analyzing the "fingerprint" spectral information of space targets is a powerful method for space target identification. In response to the needs of space target material identification and key part identification, this paper proposes a new method of computational spectral imaging with high Light utilization for space target detection. A high-resolution spatial spectral image is obtained through the combination of panchromatic channel and calculated spectral channel. Introduce the calibration technology of the system, including the target's spectrum calibration and the system's coding calibration technology. The multi-spectral image of the satellite model taken by the new spectral imaging system is used to expand the sample, and the training set data is used for training, and the entire data set is tested. The average recognition rate of the five categories of satellite main body, windsurfing board, pot body, antenna and space background is 74.86%. If only the identification of the target and the background is considered, and the non-critical part of the satellite antenna is not considered, the probability of correct recognition as a target is 98.92%, and the probability of correct recognition as a background is 99.11%.