Identifying pollutants in the internal cavities of rocket and space industry products using a multispectral endoscope

This work addresses the problem of detecting and classifying foreign contaminants on the internal surfaces of hard-to-reach cavities in rocket and space applications using a multispectral video endoscope, which simultaneously generates 16 images of the object in narrow spectral intervals across different sections of a single matrix radiation detector. The specific nature of the objects being monitored and their observation conditions dictates the spectral data processing methodology. It has been experimentally demonstrated that the reliability of monitoring is significantly improved by using a contaminant database created for a similar test object.