Underwater Image Segmentation Using Human Psycho Visual Phenomenon

Underwater Image Segmentation Using Human Psycho Visual Phenomenon

Soumyadip Dhar (RCC Institute of Information Technology, India) and Hiranmoy Roy (RCC Institute of Information Technology, India)
Copyright: © 2018 |Pages: 17
DOI: 10.4018/978-1-5225-5246-8.ch005
OnDemand PDF Download:
No Current Special Offers


In this chapter, a novel method is proposed for underwater image segmentation based on human psycho visual phenomenon (HVS). In the proposed method the texture property of an image is captured by decomposing it into frequency sub-bands using M-band wavelet packet transform. The sub-bands represent the image in different scales and orientations. The large numbers of sub-bands are pruned by an adaptive basis selection. The proper sub-bands for segmentation are selected depending on the HVS. The HVS imitates the original visual technique of a human being and it is used to divide each sub-band in Weber, De-Vries Rose, and saturation regions. A wavelet packet sub-band is selected for segmentation depending on those three regions. The performance of the proposed method is found to be superior to that of the state-of-the-art methods for underwater image segmentation on standard data set.
Chapter Preview


The basic physics of the light propagation in the water medium is different than that of the air medium. (Schettini, 2010) . The properties of the water medium cause degradation of the underwater captured images, which is different from normal images taken in air. Deep inside the water the amount of light starts decreasing and the image gets darker. Water is approximately 800 times denser than air, and this density absorbs light quickly. As the amount of light is reduced when we go deeper, colors drop off one by one depending on their wavelengths. The blue color travels the longest in the water due to its shortest wavelength, making the underwater images to be dominated essentially by blue color.

Underwater images are basically characterized by their poor visibility due to dense water. The scenes deep inside the water is not clear rather hazy. The visibility is limited by light attenuation at about twenty meters in clear water and five meters or less in turbid water. The light attenuation is caused by absorption (which removes light energy) and scattering (which changes the direction of light path). The absorption causes the reduction of light energy and scattering results the change of pats of light. These two perturbations affect the image captured process at underwater. The underwater image is blurred due to the forward scattering which is generated by randomly deviated light from an object to the camera. On the other hand, backward scattering is caused by the fraction of the light reflected by the water towards the camera before it actually reaches the objects in the scene. Due to backward scattering is the reason behind the poor contrast of the underwater images.

The two perturbations, absorption and scattering also occur due to dissolved organic matter or small observable floating particles inside water. The floating particles increase absorption and scattering of light deep inside the water. The visibility range inside the deep water can be increased with artificial lighting. The artificial lighting also suffers from scattering and absorption. In addition, the artificial light illuminates the scene in a non-uniform fashion. This non-uniform illusion produces a bright spot in the center of the image with a poorly illuminated area surrounding it. Due the reasons mentioned above the underwater image segmentation is really a challenging task.

Segmentation is the first essential and important task in low level image processing. The segmentation is a generic term for those techniques which involves taking an image and extracting information relevant to specific picture ‘segments’, such as lines, regions and objects, and their inter-relationship. It is basically a process in which an image is segmented into subsets by assigning individual pixel to particular classes. The performance of segmentation of an image is affected by the different perturbations in an image. Due to different perturbations it is very difficult to detect the true boundaries between the regions. The underwater image suffers from different perturbations as described above. So, the segmentation of an underwater image is really a difficult task and it is a subject of attraction for the research community.

Complete Chapter List

Search this Book: