Performance Analysis of Adaptive Histogram Equalization-Based Image Enhancement Schemes
Laldinpuii Colney, Sandeep Kumar, Mahak Jodwal, Manisha Bharti, Upendra Kumar Acharya · 2023
Most security cameras are designed to capture a wide field view with the aim of covering a large area, and as a result, they often have a lower resolution than other types of cameras. The images in low-light and constricted environments commonly have the drawbacks of non-uniform illumination, low contrast, and large random noise. Many security cameras are intended to operate in low-light conditions, which can also affect the quality of the images they capture. Additionally, the place where the camera is installed, the type of lens and the settings of the camera also affect the quality of the image. Image enhancement of these low-light images is crucial for analysis and investigation for security and safety purposes. In this paper, we study and compare Adaptive Histogram Equalization techniques for image contrast improvement and enhancement for surveillance footage. Conventional Histogram Equalization (HE), along with its improved versions are analyzed and compared. Four image quality assessment metrics are considered; Visual Quality, Absolute Mean Error Brightness (AMBE), NIQUE and BRISQUE.