Cepstrum adaptive plateau histogram for dark IR night vision images enhancement

摘要

This paper presents a novel enhancement algorithm for infrared (IR) images. This suggested algorithm mixes the benefits of the adaptive plateau histogram equalization (APH) and the cepstrum of homomorphic filtering enhancement (CHFE). The main idea of this approach depends on applying the APH on the IR image. Then, the resultant image is applied on the frequency domain by using the CHFE. This CHFE model is depending on the image has produced high energy that received from object. The energy received is determined by illumination radiation source and the features of reflectance of the object itself. Applying this model on the APH gives more details in the IR image and the IR night vision images show like as in the morning. The performance quality metrics for the suggested approach are entropy, average gradient, contrast, and Sobel edge magnitude. Simulation results reveal the success of the proposed approach in enhancing the quality of IR images.