Qualification: 
B-Tech
Email: 
r_sujee@cb.amrita.edu

Sujee R. currently serves as Assistant Professor at the Department of Computer Science and Engineering, School of Engineering, Amrita Vishwa Vidyapeetham, Coimbatore Campus. Her areas of research include Image Processing and Video Analytics.

Publications

Publication Type: Conference Paper

Year of Publication Title

2020

Sujee R., Shanthosh, D., and Sudharsun, L., “Fabric Defect Detection Using YOLOv2 and YOLO v3 Tiny”, in Computational Intelligence in Data Science, Cham, 2020.[Abstract]


The paper aims to classify the defects in a fabric material using deep learning and neural network methodologies. For this paper, 6 classes of defects are considered, namely, Rust, Grease, Hole, Slough, Oil Stain, and, Broken Filament. This paper has implemented both the YOLOv2 model and the YOLOv3 Tiny model separately using the same fabric data set which was collected for this research, which consists of six types of defects, and uses the convolutional weights which were pre-trained on Imagenet dataset. Observed and documented the success rate of both the model in detecting the defects in the fabric material.

More »»

2020

Sujee R. and Dr. Padmavathi S., “Fast Texture Classification using Gradient Histogram Method”, in 2020 6th International Conference on Advanced Computing and Communication Systems (ICACCS), 2020.[Abstract]


Classification refers to a physical object as being specified in a set of predefined categories. The goal in texture classification is to provide an unknown sample image for a set of known texture classes. It involves deciding which category of the texture of a painted image. Texture classification is a popular technique used in Image Processing Fields. Its applications includes classification from satellite images in types of land use, automated paint inspection for quality check, automated inspection of defects in the textile industry. In this paper texture classification is done using the gray level co-occurrence matrix. The gradient angle is used to obtain the structural component of the texture. Instead of finding the GLCM for all the angles, in this paper the GLCM of the maximum orientation is used to classify the textures. As a result the computational time and complexity have drastically reduced. Experimental analysis was performed by changing the parameters of the co-occurrence matrix and the gradient angle.

More »»

2019

Sujee R. and Sesh, V. B., “Natural Scene Classification”, in 2019 International Conference on Computer Communication and Informatics (ICCCI), 2019.[Abstract]


Humans are very proficient at perceiving natural scenes and understanding their contents. Everyday image content across the globe is rapidly increasing and there is a need for classifying these images for further research. Scene classification is a challenging task, because in some natural scenes there will be common features in images and some images may contain half indoor and half outdoor scene features. In this project we are going to classify natural scenery in images using Matlab and image processing technique such as SURF feature extraction, K-Means clustering,LDA.

More »»

2018

Sujee R. and Dr. Padmavathi S., “Pyramid-based Image Interpolation”, in 2018 International Conference on Computer Communication and Informatics (ICCCI), 2018.[Abstract]


A technique for better reconstruction of a distorted region in an image using Pyramid-based image interpolation is presented in this paper. The central idea is to construct the image pyramid, so that the apex of the pyramid has a minimized distorted region. The reconstruction of the original image starts from the apex of the pyramid. Each time, the pixels in the distorted region are recovered using bilinear interpolation technique. Since the distorted pixels in the image at the apex of the pyramid have maximal undistorted neighbors, the accuracy of reconstruction is enhanced. Thus the bottom-up approach proposed by this paper results in improved image reconstruction when compared to top-down technique due to the above mentioned higher availability of un distorted pixels in the neighborhood of a distorted pixel.

More »»

Publication Type: Conference Proceedings

Year of Publication Title

2020

Sujee R. and Thangavel, S. Kumar, “Plant Leaf Recognition Using Machine Learning Techniques”, New Trends in Computational Vision and Bio-inspired Computing: Selected works presented at the ICCVBIC 2018, Coimbatore, India. Springer International Publishing, Cham, pp. 1433–1444, 2020.[Abstract]


Leaves can be of more importance in the context of recognition. Creating a model will help in recognizing them for different applications like Medicine and Herbal analysis. The leaves has features that can be statistical based or at high level. It can include edge and any features built over pixel level. Edge identification is used for data extraction, image segmentation and data compression. In this paper Statistical features for set of leaves are identified, Leaves are classified using multi class SVM and Edges of a leaf is identified by using canny, prewitt and sobel edge recognition techniques base on various Gaussian mask. Convolution Neural Network is used to classify the given image under 14 categories. From the trial results, it is seen that canny edge identification method gives preferable outcomes over prewitt and sobel edge recognition strategies. The paper also provides directions for using Convolution Neural Network for Leaf recognition.

More »»

2017

Sujee R. and Dr. Padmavathi S., “Image enhancement through pyramid histogram matching”, 2017 International Conference on Computer Communication and Informatics (ICCCI). IEEE, Coimbatore, India, 2017.[Abstract]


Pyramids being an emerging technology in the field of image processing, this paper uses the same for enhancing images using histogram matching. It gives a detailed analysis of enhancing the images by improving their contrasts using histogram matching in the pyramid layers thus extracting the information from the images to the maximum possible. It also shows the variation in the contrast of the images when matched with different sets of images of different contrasts.

More »»

2017

Sujee R. and Kannammal, K. E., “Energy efficient adaptive clustering protocol based on genetic algorithm and genetic algorithm inter cluster communication for wireless sensor networks”, 2017 International Conference on Computer Communication and Informatics (ICCCI). IEEE, Coimbatore, India, 2017.[Abstract]


Wireless Sensor Networks (WSN) consists of independent sensors distributed across different locations, which are used to continuously record the environmental or physical conditions. The recorded data will be passed through the network from a sensor to a main location. Sensors life time is based on the battery power. Lower the battery power leads to decrease in lifetime for a sensor. It is important to increase the lifespan of sensors and also to distribute the power over WSN. Energy can be efficiently used up to some level in one of the hierarchy routing protocol that is Low Energy Adaptive Clustering Hierarchy (LEACH). This paper compares the performance of LEACH, Genetic-LEACH and Inter cluster Communication in LEACH. Here, first analyzed the basic operations involved in LEACH and then it is optimized using Genetic Algorithm (GA) to extend the life time of WSN. Finally these results are compared with LEACH which uses inter cluster communication to reach sink instead of direct communication. First set of Simulation results using MATLAB shows that the Genetic LEACH increases the living time of the WSN compared with LEACH, and also second set of results shows that the inter cluster communication in LEACH reduces energy consumption by the nodes significantly and the living period of WSN is increased compared with LEACH and with Genetic LEACH.

More »»

2015

Sujee R. and K.E., K., “Behavior of LEACH protocol in heterogeneous and homogeneous environment”, 2015 International Conference on Computer Communication and Informatics (ICCCI). IEEE, Coimbatore, India, 2015.[Abstract]


Wireless Sensor Network (WSN) technology used to sense various types of physical and environmental conditions with the availability of small and low-cost sensor nodes. Main drawback in WSN is limited battery power in the sensor nodes. It is needed to distribute the energy dissipated through WSN and also needed to maximize the lifespan of sensor nodes. Energy efficiency can be accomplished through hierarchical routing protocols. One of the fundamental protocol in this class is Low Energy Adaptive Clustering Hierarchy (LEACH). This paper gives a survey of LEACH routing protocol for WSN and compared the performance in homogeneous and heterogeneous environment. Here, first analyzed the basic distributed clustering routing protocol LEACH, which is in a homogeneous environment, then analysed with the heterogeneity concept in nodes to increase the life of WSN. Simulation results were obtained using MATLAB that shows the LEACH heterogeneous environment significantly reduces energy consumption and increases the total lifetime of the WSN than LEACH homogeneous environment.

More »»
Faculty Research Interest: