Contrast & Color Improvement Based Haze Removal Of A Underwater Image Using Fusion Technique
issue 1

Contrast & Color Improvement Based Haze Removal Of A Underwater Image Using Fusion Technique

Utkarsha Satpute1, Mitali Kamilkar2, Kiran Thakre3, Shrey Parlewar4 , Prof. Rajendra Khule5

1234B.E. Electronics Engineering, K. D. K. College Of Engineering Nagpur

5Professor, Dept. Of Electronic Engineering, K. D. K. College Of Engineering Nagpur, Maharashtra, India

Abstract Degradation of images captured under the water are mainly because of the  Scattering and absorption of light in water. This degradation results to diminished colors, low brightness and undistinguishable objects in the image. To improve the quality of such degraded images, we have proposed fusion based underwater image enhancement technique that improves  the contrast and color of underwater images. In this we are simply improving the visibility and quality of the underwater image.

Key Words: De-hazing, scattering, haze, watershed, enhancement, contrast.


scattering and absorption of light by  dust particles and water droplets in the atmosphere leads to Haze formation. Due to the haze, image loses its fidelity and contrast . To overcome the  effects of haze, de-hazing of the image is needed, which is a challenging task because the concentration of haze varies across locations making haze levels and persistence difficult to predict and address. Unclear images can reduce the effectiveness of visual monitoring in a variety of applications. These images, even if taken with high resolution cameras, are typically undermined by haze. In scenes affected by haze, de-hazing methods enhances the quality of the images to make them usable in further processing.

As the technology using digital imaging and computer vision is being used in more different areas, de-hazing is gaining more attention in research. De-hazing has a special importance in applications where the images are degraded significantly by the environment. The following figure shows the haze formation in underwater environment.

Figure 1. formation of haze

Physical properties of light during propagating through water cause the underwater scenes be affected by the absorption and scattering of light in the turbid medium. The light as an electromagnetic radiation is significantly absorbed by the water as it propagates. Moreover, based on current turbidity conditions, water may contain wide range of different solid particles that obstruct, scatter or refract incident light. As a result, the images captured inside water suffer from low contrast, blur, color distortion, and uneven attenuation. Therefore, de-hazing methods developed for underwater scenes have to deal with the problems caused by the environment differently than the de-hazing methods used in conventional scenes set in the atmosphere. Following diagram explains the scattering of the light  in the atmosphere.

Figure 2. scattering effect

In this paper we used the watershed algorithm for de-hazing the underwater image. Here we are improving  the color and the contrast of the image. The watershed algorithm is a common image segmentation technology which converts the image into topographic map to search out all the catchments in the image. The principle of watershed algorithm is to obtain the gradient image of an image before using horizontal and vertical coordinates and image gradient value to display the three-dimensional topographic map of gradient image. For the topographic map of a gradient image, the difference in the gradient value can cause fluctuant terrain, so there are many catchment basins and watersheds between catchment basins. The concept of watershed algorithm is constructed on this topographic map. The watershed algorithm recognition uses the above concept to mark the position with maximum change in image gray-scale value on the image.  


A block diagram of  the proposed methodology is shown in the figure 3. The watershed algorithm is one of the effective method for removing the haze of the underwater image. In this paper we have taken a underwater image and tried to remove the haze of the taken underwater image using watershed algorithm.


In this, we are de-hazing the given image using MATLAB code in the MATLAB using the DIP tool present in it. First it read the input image, as every digital image is in the RGB format. This is all includes in the preprocessing step. In the processing stage the RGB formatted image is get converted into the GRAY scale. After reading the image, it converted into the image matrix which is in the for of the no. of rows and the no. of columns or simply into the pixel values. As the image gets converted into the pixel values, the watershed algorithm is applied on the image. Due to the watershed algorithm, the corrected pixel values or the de-hazed image which is in the form of gray scale is now converted into the RGB scale. And hence we obtained the de-hazed image. Following is the input underwater image which is taken for the de-hazing purpose. 

Figure 3.Input Hazy Image


Applications using computer vision algorithms require certain degree of input image quality to provide successful results. Some methods are particularly sensitive to artifacts that may appear in images and videos captured in uncontrolled environment such as noise, low contrast, color distortions, or lack of light in the scene. As it is not possible to assure ideal conditions in all situations, additional processing has to be implemented to mitigate the negative weather conditions in the environment. One of such methods is de-hazing, which aims to improve the visibility of images that contain haze. These phenomena are usually present in outdoor scenes where different weather conditions greatly influence the quality of acquired digital images. Haze can be described as a situation when small solid or liquid particles are suspended in the medium where light is propagating. The light traveling through such environment interferes with these particles, and is subsequently absorbed or scattered by them. This interaction alters the light captured in the camera and results in images that are different than the real radiance of the scene. Visual effects caused by interaction of light with the particles present in the atmosphere can be divided into three categories: scattering, absorption, and emission. A ray of light hitting a particle transfers its energy to it. While a portion of the light’s energy is absorbed by the particle, the rest of the energy is non-uniformly scattered into different directions in respect to the direction of incoming light. In addition, the light scattered back to the environment further interferes with the scene. Complex interactions between particles concentrated in the atmosphere therefore result in blurring, attenuation of the light along the line of sight, lack of contrast, and color distortion. The result of the interaction between light and a particle depends on the material, shape, and size of the particle. Since the particles present in the atmosphere are mostly miniature water droplets, their shape and material is generally the same in different conditions. Varying size of the particles, however, dramatically changes the way the light is scattered by the particles. It has been observed that the scattering function of a particle is directly related to its size in respect to the wavelength of incident light. If the size of particles is less than the wavelength of the light, the scattering is more equally distributed in both forward and backward direction of the incoming light. On the other hand, the particles of size equal or greater than the wavelength of the light tend to scatter the light entirely in the forward direction, while absorbing most of the light’s energy. Since the scattered light is propagating further to the environment, each particle can also be considered a point light source emitting light into different directions according to its specific scattering function. Scattering of all particles in the atmosphere collectively creates an ambient light reflected into the scene called atmospheric light or air light, which is the source of  blur and color shifts in observed images.


The results of the proposed method are shown in the figures below

Figure a. input image
Figure b. YUV conversion
Figure c. luminance
Figure d. chrominance of U image
Figure e. chrominance of V image
Figure f. enhanced intensity image
Figure g. enhanced image
Figure h. de-haze image

Above images shows the step by step improved images according to the algorithm used .


In this paper, the image de-hazing algorithm is proposed. To improve the quality of the image we perform the image segmentation using watershed method. As a result we get enhanced and improved image free of haze.  


  1.  F. Azari Nasrabada , H. Hassanpoura , S. Asadi Amiri* Image Dehazing Improving Dark Channel Prior, IJE TRANSACTIONS B: Applications Vol. 32, No. 2, (February 2019) 284-291 ,  Faculty of Computer Engineering and IT, Shahrood University of Technology, Shahrood, Iran and Department of Technology and Engineering, University of Mazandaran, Babolsar, IranF. Azari Nasrabad et
  2. M.Suganthy*, S. Lakshmi, S.Palanivel, Enhancing the Quality of Underwater Images using Fusion of  sequential Filters and Dehazing  in  International Journal of Engineering & Technology, 7 (2.24) (2018) 296-298 
  3. Codruta O. Ancuti , Cosmin Ancuti, Christophe De Vleeschouwer , and Philippe Bekaert, JANUARY 2018 ‘Color Balance and Fusion for Underwater Image Enhancement’ IEEE
  4. An Adaptive Image Dehazing Algorithm based on Dark Channel Prior in 2017 29th Chinese Control And Decision Conference (CCDC) 978-1-5090-4657-7/17/$31.00_c 2017 IEEE by Chunlin Chen1, Jiatong Li1, Sibin Deng1, Feng Li1, Qiang Ling1,. University of Science and Technology of China, Hefei, 230022, China
  5. A Novel Dark-Channel Dehazing Algorithm Based on Adaptive-Filter Enhanced SSR Theory DOI: 10.4236/jcc.2017.511005 Sep. 25, 2017
  6. T.Padmapriya and V.Saminadan, “Utility based Vertical Handoff Decision Model for LTE-A networks”, International Journal of Computer Science and Information Security, ISSN 1947-5500, vol.14, no.11, November 2016.
  7. S.V.Manikanthan and K.srividhya “An Android based secure access control using ARM and cloud computing”, Published in: Electronics and Communication Systems (ICECS), 2015 2nd International Conference on 26-27 Feb. 2015, Publisher: IEEE,DOI: 10.1109/ECS.2015.7124833.
  8. M. Suganthy ,P. Ramamoorthy and A. Kanagalakshmi, 2014, ‘Wavelet And Texture Feature Fusion Of Dual IrisBased Security System Using T-Norm’ vol. 9, no. 24 ,pp. 30681-30692 Transactions On Image Processing, VOL. 27, NO. 1, pp-379-393
  9. CodrutaOrnianaAncuti and Cosmin Ancuti, (2013)“Single Image Dehazing By Multi-Scale Fusion” IEEE Transactions On Image Processing, vol. 22, no. 8.
  10. M.Suganthy,P.Ramamoorthy , 2012, ‘An Efficient authentication By Iris Using Log Gabor Filter and Neural network’ , IJCSI International Journal of Computer Science Issues, vol. 9, no.5, pp.1694-0814
  11. Prabhakar C.J.1*, Praveen Kumar P.U. ‘An Image Based Technique For Enhancement Of Underwater Images ‘International Journal of Machine Intelligence ISSN: 0975–2927 & E-ISSN: 0975–9166, Volume 3, Issue 4, 2011, pp-217-224.
  12. Raimondo S, Silvia C. 2010,’Underwater image processing: state of the art of restoration and image enhancement methods’ [J]. EURASIP Journal on Advances in Signal Processing.
  13. Petit F, Capelle-Laize A, Carré P. 2009,’ Underwater image enhancement by attenuation inversionwith quaternions[C]’, Acoustics, Speech and Signal Processing, 2009. ICASSP 2009. IEEE International Conference on. IEEE,: 1177-1180
  14. Iqbal K, Abdul Salam R, Osman M, et al., 2007, ‘ Underwater Image Enhancement Using An Integrated Colour Model[J]’, IAENG International Journal of Computer Science, , vol.32, no.2 pp.239-244.
  15. Schechner Y Y, Karpel N, 2005, ’ Recovery of underwater visibility and structure by polarization analysis[J]’, Oceanic Engineering, IEEE Journal , vol.30, no.3, pp.570-587.
  16. Schechner Y Y, Averbuch Y, 2007, ‘ Regularized image recovery in scattering media[J]’. Pattern Analysis and Machine Intelligence, IEEE Transactions on, vol. 29,no.9, pp. 1655-1660.
  17. Kocak D M, Caimi F M. The current art of underwater imaging– With a glimpse of the past and vision of the future[J]. Marine Technology Society Journal, 2005, vol.39, no.3, pp. 5-26.
  18. Schechner Y Y, Karpel N, 2004, ‘ Clear underwater vision[C]//Computer Vision and Pattern Recognition’, CVPR 2004. Proceedings of the IEEE Computer Society Conference on. IEEE, 2, vol. 1,no.1, pp. I-536-I-543.
  19. He D M, Seet G G L, 2004, ‘Divergent-beam lidar imaging in turbid water[J]’, . Optics and lasers in engineering, vol.41, no.1, pp.217-23
  20.  Narasimhan S G, Nayar S K, 2003, ‘ Contrast restoration of weather degraded images[J]’, Pattern Analysis and Machine Intelligence, IEEE Transactions on, vol.25, no.6, pp. 713-724.

Related posts

Blood and organ Managment system


Scaling and analysing Happiness Index Using Data Science


MetaBOT Android Application


Leave a Comment