Indian Journal of Science and Technology

Abstract

Background/Objectives: Contrast enhancement is an important stage in uncooled thermal images as uncooled thermal images are low-contrast images. Histogram equalization is the widely used technique for contrast enhancement of images. Methods/Statistical Analysis: The proposed method for contrast enhancement is based on plateau histogram equalisation. It preserves all the benefits of classical histogram equalisation but does not extend the contrast of background enormously24. Plateau histogram equalization technique is one of the contrast enhancement techniques where the pels count is limited to an inception value defined based on histogram properties. Findings: In comparing the results of proposed contrast enhancement system, the image has been significantly enhanced without enhancing the background gray levels and the noise and thereby increasing the brightness preservation and harvested a more natural enhancement. Applications/ Improvements: The proposed technique on thermal images provides a good enhancement, which is difficult to achieve with traditional Histogram equalization technique on thermal images. The proposed algorithm is tested on raw images, gain corrected images, and Gain and Offset corrected images and achieved satisfactory results.
Keywords: Contrast Enhancement, Histogram Equalization, Low-contrast images, Plateau Histogram Equalization, Thermal Images