Scientific Publications
Browse our catalog of peer-reviewed scientific articles
Research Categories
Select a discipline to explore published research papers
Computer Science & Engineering
Core computing paradigms, algorithm design, software architectures, advanced computing theory, and hardware engineering.
Computer Networks
Network architectures, wireless ad-hoc protocols, vehicular networks (VANET), buffering, and communications.
Electronics Department
Circuits, semiconductor devices, flexible electronics, solar array modeling, signal processing, and photonics.
Civil Engg
Structural analysis, thermal stability of fluid layers, environmental infrastructure, and applied materials engineering.
Recent Publications (2)
An Optic Disc Segmentation in Retinal Images using Efficient Clustering method
Optical disc segmentation is essential stage for retinal image analysis and disease identification. Segmentation locates the optic disk in the retinal fundus images. Various researchers have triad for improving the efficiency of the optical disc segmentation. Pprincipal component analysis based method is popularly used but are computationally complex. Thresholding is also widely used for improving the efficiency of the disc detection method. But under diabetic patient there efficiency degrades. Fuzz C means clustering is also used by some researchers but the efficiency is the major concern under optical imaging environment. This paper presents a new method of optical disc segmentation using the K mean clustering in retinal images. Study of thresholding based methods for optical disc segmentation is presented. It is observed that the method of disc detection should be less complex and efficiently works for all kind of images.
EDGE DETECTORS: SURVEY
Image segmentation is the technique of detecting edges of the object within the picture that we want to identify. Image recognition is carried out in various steps: Step 1. RGB value of picture is extracted. Step 2. Gray image is acquired either using formulae 0.2989 * R + 0.5870 * G + 0.1140 * B or the usage of rgb2gray () function. Step 3. Binary image is acquired from gray image. Step 4. Now edge detection is implemented using various algorithms along with Sobel, Canny, Roberts, Prewitt, Laplacian of Gaussian, Zero cross detection. Edge detection is the most vital procedure as it symbolizes how nicely the object could be detected. This research papers gives comparison on existing methods such as Sobel operator, canny operator, Prewitt operator, Roberts operator, zero crossing method, Laplacian of Gaussian method using MATLAB R2016a.