Robust air-light estimation for a single hazy image using haze-lines in Plüker coordinates

Area of Science:

• Computer Vision
• Image Processing

Background:

• Image dehazing is crucial for mitigating adverse weather effects on outdoor optical imaging.
• Accurate air-light estimation is fundamental for effective image dehazing.
• Current methods often fail due to bright objects like lamps, impacting air-light estimation.

Purpose of the Study:

• To propose a novel method for robust and optimal air-light estimation.
• To address the limitations of existing air-light estimation techniques.
• To enhance the accuracy of image dehazing algorithms.

Main Methods:

• A new method for air-light estimation using haze-lines in Plüker coordinates is proposed.
• The analytic solution for air-light is derived by analyzing geometric relationships in 3D RGB space.
• A novel algorithm is developed based on haze-lines in Plüker coordinates.

Main Results:

• The proposed algorithm demonstrates more robust and accurate air-light estimation compared to existing methods.
• Experimental validation confirms the effectiveness of the novel approach.
• The method successfully overcomes limitations caused by bright objects in hazy images.

Conclusions:

• The proposed haze-line-based air-light estimation method provides a significant improvement in image dehazing.
• This technique offers a more reliable solution for challenging outdoor imaging conditions.
• The findings contribute to advancing the field of image processing and computer vision.