Bilkent University
Department of Computer Engineering

---

CS 590/690 SEMINAR

 

Power-Efficient Low Light Image Enhancement Using Adaptive Gamma Correction and FPGA Acceleration

 

İlayda Sarıçam
Master Student
(Supervisor: Prof.Dr.Uğur Güdükbay)
Computer Engineering Department
Bilkent University

Abstract: In recent years, processing low-light images has become a significant challenge due to the need for enhanced image quality in various applications. To solve this problem, both traditional and training-based approaches have been explored. This research focuses on enhancing the visualization of dark images by using a traditional method with a novel feature, particularly for tasks that require low power consumption. Gamma correction approach, a well-known traditional technique, is used to enhance image quality. However, instead of applying a constant gamma value across the entire image, an optimal gamma value is selected for each region. To determine the optimal gamma value, an edge detection algorithm is used to segment the image into regions based on their local characteristics such as intensity and dynamic range. These regions are then assigned specific gamma values, improving the overall image quality. Moreover, to optimize processing speed, the computationally intensive tasks of edge detection and gamma correction are implemented on an FPGA, leveraging its parallel processing capabilities. Also, the algorithm is tested in LOL dataset, which includes both dark and light images, to compare the effectiveness of our approach with performance metrics such as signal-to-noise ratio (PSNR) and the structural similarity index measure (SSIM).

 

DATE: December 2, Monday @ 16:10 Place: EA 502