Effective method for low-light image enhancement based on the JND and OCTM models

Area of Science:

• Computer Vision
• Image Processing
• Digital Signal Processing

Background:

• Low-light images exhibit degraded quality due to dim brightness, low contrast, and limited dynamic range.
• Existing image enhancement techniques often struggle with detail preservation and overall quality.

Purpose of the Study:

• To propose an effective method for low-light image enhancement.
• To improve brightness, contrast, and detail preservation in degraded images.
• To outperform current state-of-the-art low-light enhancement algorithms.

Main Methods:

• Image decomposition into base and detail layers using a guided filter.
• Detail enhancement via a visual masking model.
• Brightness adjustment of base images using just-noticeable-difference (JND) and optimal contrast-tone mapping (OCTM) models.
• Generation of artificial images for refined brightness adjustment and detail preservation.

Main Results:

• Successfully enhanced low-light images, improving brightness and contrast.
• Achieved superior detail preservation compared to single-input algorithms.
• Demonstrated qualitative and quantitative improvements over state-of-the-art methods.

Conclusions:

• The proposed method effectively enhances low-light images.
• The integration of JND and OCTM models, along with a novel artificial image generation technique, yields superior results.
• The approach offers a significant advancement in low-light image enhancement technology.