Haze Removal via NIR-Assisted RGB Image Decomposition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for removing haze from RGB images often result in noise and bluish artifacts, especially when most pixel data are lost due to dense haze, failing to effectively restore the image outline.
Innovation Solution
A device comprising an image decomposer, a weight generator, a detail layer mixer, a base layer dehazer, and an adder, which decomposes NIR and RGB images, generates mixing weight values based on high-frequency component similarities, mixes detail layers, and compensates base layers to produce an output image free of haze.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If conventional haze removal schemes are used on RGB images, then haze can be removed to some extent, but noise and bluish artifacts are introduced and the image outline is lost
Solution Approach 1:
The patent segments the image processing task into two distinct parts: base layer processing for haze removal and detail layer processing for outline preservation. The RGB image is decomposed into base and detail layers, where the base layer undergoes haze removal processing while the detail layer retains high-frequency information. This segmentation allows independent optimization of each layer's processing, preventing the loss of image outlines that occurs in conventional unified processing approaches.
Solution Approach 2:
The patent introduces an intermediary NIR (near-infrared) image to assist in the haze removal process. The NIR image, which penetrates haze more effectively, serves as a mediator to guide the restoration of the RGB image's detail layer. By using the NIR image as an intermediary reference, the system can recover lost high-frequency information and image outlines without introducing noise or artifacts directly from conventional haze removal methods.
2Object-affected harmful factors
If conventional haze removal schemes are used on RGB images, then haze can be removed, but noise and bluish artifacts are caused
Solution Approach 1:
The patent segments the image processing task into two distinct parts: base layer processing for haze removal and detail layer processing for outline preservation. The RGB image is decomposed into base and detail layers, where the base layer undergoes haze removal processing while the detail layer retains high-frequency information. This segmentation allows independent optimization of each layer's processing, preventing the loss of image outlines that occurs in conventional unified processing approaches.
Solution Approach 2:
The patent introduces an intermediary NIR (near-infrared) image to assist in the haze removal process. The NIR image, which penetrates haze more effectively, serves as a mediator to guide the restoration of the RGB image's detail layer. By using the NIR image as an intermediary reference, the system can recover lost high-frequency information and image outlines without introducing noise or artifacts directly from conventional haze removal methods.
Data Source
AI summary
A device for removing haze from an image includes an image decomposer that decomposes a near-infrared (NIR) image to generate an NIR detail layer image and decomposes an RGB image to generate an RGB detail layer image and an RGB base layer image, a weight generator that generates a mixing weight value based on a similarity between high frequency (HF) components of the NIR image and the RGB image, a detail layer mixer that mixes the NIR detail layer image and the RGB detail layer image based on the mixing weight value to generate a mixed RGB detail layer image, a base layer dehazer that removes haze from the RGB base layer image to generate a compensated RGB base layer image, and an adder that adds the mixed RGB detail layer image and the compensated RGB base layer image to generate an output RGB image.