An approach for underwater image enhancement based on color correction and dehazing

This paper proposed an improved approach to enhance the quality of underwater images without the aid of any specialized hardware. The proposed method consists of three steps including color correction, contrast synthesis and dehazing. The color correction removes the color cast problem, contrast synthesis removes under-exposure problem and dehazing removes the fuzz problem. In the proposed method, color correction, contrast synthesis and dehazing are developed based on a statistical method, Retinex-model and utilizing the dark channel prior information respectively. After removing these three difficulties the quality of the enhanced underwater images is compared with the baseline approaches based on the value of chroma, contrast and saturation. The proposed method obtains Underwater Color Image Quality Evaluation (UCIQE)value 0.66574 which is the best among the methods compared.

Underwater images hold immense value for various fields, including marine biology research, underwater infrastructure inspection, and exploration activities. However, capturing high-quality images underwater proves challenging due to light absorption and scattering leading to color distortion, blue green hues. Additionally, these phenomena decrease contrast and visibility, hindering the ability to extract valuable information. Existing image enhancement methods often struggle to achieve accurate color correction while preserving crucial image details. This article proposes a novel deep learning-based approach for underwater image enhancement that leverages the power of autoencoders. Specifically, a convolutional autoencoder is trained to learn a mapping from the distorted colors present in underwater images to their true, color-corrected counterparts. The proposed model is trained and tested using the Enhancing Underwater Visual Perception (EUVP) and Underwater Image Enhancement Benchmark (UIEB) datasets. The performance of the model is evaluated and compared with various traditional and deep learning based image enhancement techniques using the quality measures structural similarity index (SSIM), peak signal-to-noise ratio (PSNR) and mean squared error (MSE). This research aims to address the critical limitations of current techniques by offering a superior method for underwater image enhancement by improving color fidelity and better information extraction capabilities for various applications. Our proposed color correction model based on encoder decoder network achieves higher SSIM and PSNR values.

Twice Mixing: A rank learning based quality assessment approach for underwater image enhancement

Underwater images have been drawing sufficient attentions, as more and more researchers and companies are committed to ocean engineering, underwater archaeology, mining etc in recent years. Underwater images often suffer from unbalanced illumination, which bring image analysis and processing challenges. This paper introduces a novel approach for underwater image enhancement based on the DCT coefficient and CLAHE algorithm. Image changes with operations on the statistical model of the image DCT coefficients. The proposed methods firstly perform brightness equalization of the image, which optimize the underexposed and overexposed regions by regulating the low-frequency part of the image DCT coefficients. Afterwards CLAHE is applied to achieve image contrast enhancement. Finally experiments shows the better image enhancement performance compared to other algorithms.

In this paper, a novel underwater image enhancement approach was proposed. This approach includes use of a method formed by the wavelet transform and the differential evolution algorithm. In the method, the contrast adjustment function was applied to the original underwater image first. Then, the homomorphic filtering technique was used to normalize the brightness in the image. After these steps, the underwater image was separated into its R, G, and B components. Then wavelet transform function was performed on each of the R, G, and B channels with Haar wavelet decomposition. Thus, detailed images were obtained for each of the color channels by wavelet transform low-pass approximation (cA), horizontal (cH), vertical (cV) and diagonal (cD) coefficients. Four parameters of weights (w) of each component cA, cH, cV, and cD situated in the R, G, and B color channels were optimized using differential evolution algorithm. In the proposed method, differential evolution algorithm was employed to find the optimum w parameters for Entropy and PSNR in separate approaches. Finally, unsharp mask filter was used to enhance the edges in the image. As an evaluation approach, performance of the proposed method was tested by using the criteria of entropy, PSNR, and MSE. The obtained results showed that the effectiveness of the proposed method was better than the existing techniques. Likewise, the visual quality of the image was also improved more thanks to the proposed method.

Underwater photography is important for the works of exploring underwater scenes. Comparing with normal camera images, the underwater images usually have low brightness-contrast and with low visual quality. This paper proposes a new method to improve the underwater image visual quality by using a two-step approach. Firstly, a transmission map based enhancement is applied to increase the image global contrast, similar to the image defog algorithm. Secondly, image details are extracted and the local contrast is improved by applying edge-preserving filter. Experiments demonstrated that the proposed two-step approach can significantly improve the visual quality of underwater images.

An RGB YCbCr Processing method (RYPro) is proposed for underwater images commonly suffering from low contrast and poor color quality. The degradation in image quality may be attributed to absorption and backscattering of light by suspended underwater particles. Moreover, as the depth increases, different colors are absorbed by the surrounding medium depending on the wavelengths. In particular, blue/green color is dominant in the underwater ambience which is known as color cast. For further processing of the image, enhancement remains an essential preprocessing operation. Color equalization is a widely adopted approach for underwater image enhancement. Traditional methods normally involve blind color equalization for enhancing the image under test. In the present work, processing sequence of the proposed method includes noise removal using linear and non-linear filters followed by adaptive contrast correction in the RGB and YCbCr color planes. Performance of the proposed method is evaluated and compared with three golden methods, namely, Gray World (GW), White Patch (WP), Adobe Photoshop Equalization (APE) and a recently developed method entitled “Unsupervised Color Correction Method (UCM)”. In view of its simplicity and computational ease, the proposed method is recommended for real-time applications. Suitability of the proposed method is validated by real-time implementation during the testing of the Autonomous Underwater Vehicle (AUV-150) developed indigenously by CSIR-CMERI.

Efficient underwater visual environment perception is the key to realizing the autonomous operation of underwater robots. Because of the complex and diverse underwater environment, the underwater images not only have different degrees of color cast but also produce a lot of noise. Due to the existence of noise in the underwater image and the blocking effect in the process of enhancing the image, the enhanced underwater image is still rough. Therefore, an underwater color-cast image enhancement method based on noise suppression and block effect elimination is proposed in this paper. Firstly, an automatic white balance algorithm for brightness and color balance is designed to correct the color deviation of underwater images and effectively restore the brightness and color of underwater images. Secondly, aiming at the problem of a large amount of noise in underwater images, a noise suppression algorithm for heat conduction matrix in the wavelet domain is proposed, which suppresses image noise and improves the contrast and edge detail information of underwater images. Thirdly, for the block effect existing in the process of enhancing the underwater color-cast image, a block effect elimination algorithm based on compressed domain boundary average is proposed, which eliminates the block effect in the enhancement process and balances the bright area and dark area in the image. Lastly, multi-scale image fusion is performed on the images after color correction, noise suppression, and block effect elimination, and finally, the underwater enhanced image with rich features is obtained. The results show that the proposed method is superior to other algorithms in color correction, contrast, and visibility. It also shows that the proposed method corrects the underwater color-cast image to a certain extent and effectively suppresses the noise and block effect of the underwater image, which provides theoretical support for underwater visual environment perception technology.

Underwater images are affected by factors such as haze, insufficient light, high concentration of various impurities which leads to a blue/green tinge, and blurriness. Majority of image enhancement techniques have been developed for atmospheric images but comparatively, very few techniques have been proposed for underwater image enhancement. This paper proposes a fusion algorithm which can enhance underwater images shot under various environmental conditions. This method uses Multi-Scale Retinex with Color Restoration (MSRCR) to obtain an image with improved color retention and better dynamic range. Then, the image blur is removed by using Dark Channel Prior (DCP). The results show that the proposed method gives 10% better Structural Similarity Index (SSIM) for the fusion-based algorithm.

Recently, the domain of computer vision has experienced tremendous study and experimentation. Multiple machine learning (ML) and/or deep learning (DL) approaches are inculcated or proposed to address issues in computer vision tasks like image or video classification, object detection, image enhancement, etc. We present one such technique for enhancing underwater images. This experiment tweaks the earlier proposed FUnIE-GAN architecture to conducive images with a better visual perception. The UIEB data set is utilized for paired training. There is an attempt to apply U-Net architecture as a discriminator in place of the previously adopted Markovian patch GAN. The evaluation of the model uses PSNR and SSIM (full-reference evaluation metrics), along with entropy and UCIQE (non-reference evaluation metrics). The mean values of PSNR, SSIM, UCIQE and entropy metrics for 90 test images are 17.248, 0.6, 0.513 and 6.093, respectively, indicating that the improved FUnIE-GAN version generated underwater images with better visual perception.

An underwater image often suffers from quality degradation issues, such as color deviations, low contrast, and blurred details, due to the absorption and scattering of light. In this article, we propose to address the aforementioned degradation issues via attenuated color channel correction and detail preserved contrast enhancement. Concretely, we first propose an underwater image color correction method. Considering the differences between superior and inferior color channels of an underwater image, the inferior color channels are compensated via especially designed attenuation matrices. We then employ a dual-histogram-based iterative threshold method and a limited histogram method with Rayleigh distribution to improve the global and local contrast of the color-corrected image, thus achieving a global contrast-enhanced version and a local contrast-enhanced version, respectively. To integrate the complementary merits between the global contrast-enhanced version and the local contrast-enhanced version, we adopt a multiscale fusion strategy to fuse them. Finally, we propose a multiscale unsharp masking strategy to further sharpen the fused image for better visual quality. Extensive experiments on four underwater image enhancement benchmark data sets demonstrate that our method effectively enhances underwater images qualitatively and quantitatively. Besides, our method also generalizes well to the enhancement of low-light images and hazy images.

Multi-core token mixer: a novel approach for underwater image enhancement

RASS-U-Net: a hybrid approach for underwater image enhancement with multi-scale feature extraction and attention mechanisms

Underwater color image processing has received considerable attention in the last few decades for underwater image-based observation. In this article, a novel underwater image enhancement approach using combining schemes is presented. This study aims to improve color correction under nonuniform illumination conditions. The objective of this approach is threefold. First, to correct nonuniform illumination and enhance contrast in the image, homomorphic filtering is used. Second, the color contrast of an image is equalized by a contrast stretching algorithm in RGB (red, green and blue) color space. Finally, the noise amplified after the previous two steps is suppressed by using wavelet domain denoising based on threshold processing. The comparison of experimental results shows that the proposed approach of underwater image enhancement can correct the color imbalance and is especially suitable for processing underwater color images that have nonuniform lighting.

Color correction and restoration based on multi-scale recursive network for underwater optical image