Volume 14, No 2, 2017
Convex Programming Approach For Unified Image Enhancement And White Balancing
M. Jeyakarthic
Abstract
This paper introduces a novel Convex Programming Approach for Unified Image Enhancement and White Balancing, presenting a comprehensive framework for addressing various image enhancement tasks. Leveraging an analysis of the intricate relationships between image histograms and the processes of contrast enhancement and white balancing, the proposed model is a generalized equalization model that integrates tasks into a unified convex programming framework. The model offers flexibility through parameter configurations, enabling the accomplishment of diverse enhancement goals. By defining histogram transform properties such as contrast gain and nonlinearity, optimized model parameters for specific enhancement applications. Additionally, derived an optimal image enhancement algorithm that balances contrast enhancement and white balancing, achieving a trade-off between contrast improvement and tonal distortion. Through subjective and objective experimental evaluations, demonstrates the effectiveness of our proposed algorithm in tasks such as image enhancement, white balancing, and tone correction. Furthermore, the computational complexity of our method, highlighting its practical feasibility was analysed. Overall, our Convex Programming Approach offers a versatile and efficient solution for unified image enhancement and white balancing.
Pages: 198-209
Keywords: Image enhancement, White balancing, Convex programming, Histogram transform, Contrast enhancement.