Improving image luminosity and contrast variation using hybrid statistical strategy

Abstract

Luminosity and contrast variation problems are among the most challenging tasks in the image processing field especially to improve the image quality. Enhancement is implemented by performing an adjustment of the dark or bright intensity in order to improve the quality of the images and to increase the segmentation performance. Recently, numerous methods had been proposed to normalize the luminosity and contrast variation. In this study, a new method based on a direct technique using a statistical data that is known as Hybrid Statistical Enhancement (HSE) is proposed. The HSE method used the mean and standard deviation of a local and global neighbourhood and classified the pixel into three groups; the foreground, border, and problematic region (contrast & luminosity). Two datasets namely document image and weld defect image were utilized to demonstrate the effectiveness of the HSE method. The results from the visual and objective aspects showed that the HSE method can normalize the luminosity and enhance the contrast variation problem effectively, compared to the other enhancement methods such as Homomorphic Filter and Discrete Cosine Transforms (DCT). Then, the segmentation process was done using the resulting image from the HSE method. In order to prove the HSE effectiveness, a few image quality assessments were presented and the results were discussed. The HSE method achieved the highest result compared to the other methods which are (Signal Noise Ratio = 9.32) for document dataset and (Signal Noise Ratio = 8.92) for weld defect dataset. In segmentation stage, the Otsu method obtained the highest average increment, which is 41% for document dataset and 82% for weld defect dataset. In conclusion, the implementation of the HSE method has produced an effective and efficient result for background correction, quality images improvement and increase the quality of segmentation result in term of Accuracy and Peak Signal Noise Ratio (PSNR).