Abstract
We propose a method to identify image noise type for an automatic target recognition system. In previous studies, kurtosis and skewness of image noise have been considered during identification. However, these two features vary according to each image, whereby the identification accuracy is not convincing. In order to maintain the performance of noise identification according to various images and intensities, we carried out a logistic regression analysis and designed a model-based image noise identification method using random sample consensus (RANSAC). It was confirmed that the proposed algorithm identifies 3 types of image noise according to 50 different images and 4 different noise levels.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
