Search in Medwell
 
 
Journal of Engineering and Applied Sciences
Year: 2018 | Volume: 13 | Issue: 11 | Page No.: 3916-3921
DOI: 10.36478/jeasci.2018.3916.3921  
Quality Measures Technique for Underwater Images Using Artificial Neural Network
Priya Sahotra and Simrandeep Singh
 
Abstract: The visibility in underwater images is usually bad because of the attenuation of light in water that causes the dizzy contrast and the color variation. In this proposed research an innovative and productive approach for underwater image quality enhancement will have presented. The proposed method intends to prepare better underwater image contrast, increase image details with fewer losses in information and reduce noise by applying an innovative procedure of using contrast stretching to produce two different images with different contrasts. Capturing images through digital camera or smart phones has become simple nowadays. Because of the enhancement of camera properties, image in underwater and low light conditions is suffered from exposure problem. The proposed method is combination of different methods which is successful in magnify the capture in underwater by using the artificial neural network. Underwater image quality measures technique for underwater images using artificial neural network will have proposed. The underwater image quality measures technique is proposed for the image quality improvement on the basis of three different techniques. For the recognition of suitable pixel sets from the underwater image we use image colorfulness, sharpness and contrast based pixels and train a proposed system to classify that pixel set by using Artificial Neural Network (ANN). At last performance metrics of proposed underwater image quality measures is calculated and compared with previous existing research.
 
How to cite this article:
Priya Sahotra and Simrandeep Singh, 2018. Quality Measures Technique for Underwater Images Using Artificial Neural Network. Journal of Engineering and Applied Sciences, 13: 3916-3921.
DOI: 10.36478/jeasci.2018.3916.3921
URL: http://medwelljournals.com/abstract/?doi=jeasci.2018.3916.3921