A modified Histogram Equalization approach for Image Contrast Enhancement

• Published in: 2022 2nd International Conference on Consumer Electronics and Computer Engineering (ICCECE) pp. 545 - 550
• Main Authors: Li, Yang, Sun, Runmin, Luo, Chen, Zhou, Yijun
• Format: Conference Proceeding
• Language: English
• Published: IEEE 14.01.2022
• Subjects: Computational efficiency; Contrast Enhancement; Face recognition; Gray-Level Distribution; Gray-Level Peak; Histogram Equalization; Histograms; Illumination; Image segmentation; Lighting; Local Minimum; Real-time systems; Segmentation Point; Streaming media
• DOI: 10.1109/ICCECE54139.2022.9712841

Histogram equalization is a widely used technique for adjusting image intensities to enhance contrast. However, output images often experience a loss of fine details or contain unwanted viewing artefacts from over-enhancement. Existing modified methods tend to weaken enhancement effect due to multiple segmentation. Moreover, the implementation of these modified methods is generally hampered by newly introduced computation complexity. In this paper, a new image contrast enhancement method has been proposed. It divides the histogram into two parts and performs histogram equalization for each sub-histogram separately. The dividing procedure is applied only once to improve enhancing effect while reduce processing time. The selection of segmentation point is based upon close examination of characteristics of histogram to ensure proper pixel distribution. Experiment results show the developed technique outperforms existing modified ones. The advantage is more pronounced on handling images captured under poor illumination conditions.