An Automatic Method to detect defects for Solid Rocket Motor

Since the traditional defects detection methods on the CT image was not highly automated and could not lead to satisfied effects, a new defects detection method of the solid rocket motor was proposed based on the bot-hat transformation and threshold segmentation. Firstly, the bot-hat transformation was adopted to transform original images to gray images with enhanced valley values; secondly, an automatic threshold segmentation algorithm was designed to segment the defects in the gray images. It was found that with the new proposed defects detection method, the spatial distributions, outlines and size characteristics of the defects could be clearly detected, which demonstrated an effective detection result.