Optical imaging remote sensing technology is an important technical means to obtain information of ground objects, but it is restricted by bad weather such as clouds, rain, and haze. In haze weather condition, optical images often have poor contrast and blurred details, which has a great impact on subsequent applications and interpretation. If the acquired images are processed, not only their quality can be improved, but also their visual effect and utilization value can be improved, so as to reduce the impact brought by haze. In general, the haze is removed from the view of image processing. In this paper, through in-depth analysis and study of existing algorithms and characteristics of optical remote sensing images, a new idea is proposed to solve this problem from the view of combination of image content and auxiliary information. Furthermore, a new haze removal algorithm is proposed based on the Retinex multi-scale model and the histogram characteristics of remote sensing images. Because the new method combines multi-scale model (MSM) and histogram characteristics (HC), it is referred to as MSMHC algorithm in this paper. The advantage of the new method is that the content and type of image are considered in the whole process, and then two processing schemes are set for haze removal. In the test experiments, one hundred groups of image data were used to carry out comparative experiments. At the same time, single-scale Retinex (SSR) algorithm, multi-scale Retinex (MSR) algorithm, dark channel priori (DCP) method, brightness preserving dynamic fuzzy histogram equalization (BPDFHE) algorithm, histogram equalization (HE) method, and homomorphic filter (HF) algorithm were used for comparative experiments with the MSMHC method. Five parameters, including standard deviation (SD), information entropy (IE), peak signal to noise ratio (PSNR), structural similarity (SSIM), and image contrast (IC), were used to quantitatively evaluate the test results. The experimental results and the parameter values showed that the MSMHC algorithm could not only effectively remove haze from remote sensing images, obtain high contrast and high definition images, but also have better generalization ability.

/pdf/haze-removal-algorithm-for-optical-remote-sensing-image-4y4rcpvhf7.pdf

Haze Removal Algorithm for Optical Remote Sensing Image Based on Multi-Scale Model and Histogram Characteristic

The enhancement of catenary image with low visibility based on multi-feature fusion network in railway industry

We present white balancing, quad-tree decomposition, and L 1 minimization-based dehazing technique. Atmospheric light is estimated through quad-tree decomposition while holistic transmission map is estimated through coarse-scale net. Coarse and fine layers are refined through fusion and a luminance model is used to preserve details and edges. The final dehazed image is refined using L 1 minimization. Simulation results when equated with existing techniques, verify the effectiveness of the proposed technique.

White balancing, quad-tree decomposition, and L1 minimization-based single-image dehazing

To address blur and noise issues caused by airlight and atmospheric scattering in nighttime imaging environment, we proposed a multi-frame image restoration method. First, the Luojia 1–01 night-lig...

Lirong Li

Papers

Instant haze removal from a single image