Qualification: 
M.Tech, B-Tech
vinithapanicker@am.amrita.edu

Vinitha J. Panicker currently serves as an Assistant Professor (Sr. Gr.) at the Department of Information Technology at Amrita School of Engineering, Amritapuri. She has completed M. Tech. in Computer Science and Engineering with specialization in Digital Image Computing with First Rank from Kerala University. She has 12 years and 8 months of academic experience.

Publications

Publication Type: Conference Paper

Year of Publication Title

2018

H. Sathyan and Vinitha Panicker J, “Lung Nodule Classification Using Deep ConvNets on CT Images”, in 2018 9th International Conference on Computing, Communication and Networking Technologies (ICCCNT), Bangalore, India, 2018.[Abstract]


Lung cancer is a hazardous disease which is the unrestricted growth of abnormal cells that can occur in one or both of the lungs. Lung tumors can be of two types such as benign and malignant. The survival rate of lung cancer depends upon early identification of lung nodules which is a crucial process. We propose a new method for Lung nodule classification from Lung CT images by using deep convolutional neural networks [19]. This method eliminates the need of manual feature extraction which is a feed back of previous works. The network is fed with raw lung CT images from publicly available LIDC-IDRI dataset. Here, the lung images are classified into three classes such as: non-nodules, nodules of size <;3 mm and nodules of size >= 3 mm. This classification is achieved with the help of AlexNet which is a pre-trained convolutional neural network with the help of transfer learning methodology. This method successfully classified the lung CT images into three classes and achieved 98% accuracy with comparatively less false positive rates.

More »»

2017

D. Dath and Vinitha Panicker J, “Enhancing adaptive huffman coding through word by word compression for textual data”, in 2017 International Conference on Communication and Signal Processing (ICCSP), Chennai, India, 2017.[Abstract]


Nowadays, the volume of information that is being processed is increasing exponentially. And hence, the significance of data compression algorithms is also increasing. Data compression algorithms aim at reducing the size of data at the cost of increased computational efforts. In this paper, we propose an enhanced version of adaptive Huffman encoding technique for improving the efficiency of text file compression. In the proposed encoding technique, we devise a mechanism for compressing the text files word by word rather than the existing byte by byte compression technique. The proposed method is implemented and the results showed better performance in terms of compression ratio as well as execution time.

More »»

2017

A. Krishnan, Jayadevan, P., and Vinitha Panicker J, “Shadow removal from single image using color invariant method”, in 2017 International Conference on Communication and Signal Processing (ICCSP), Chennai, India, 2017.[Abstract]


Shadows are caused when light from a source of light is blocked by opaque objects. In the field of image processing, the shadow cause many technical difficulties. Shadows may give us a wrong interpretation of the shape, colour, or orientation of the image. Objects may appear to merge, when shadow of one object falls on another object and making the object count lesser. Hence removal of shadows is a very crucial and inevitable task of many of computer vision algorithms, such as segmentation, object detection and tracking etc. Shadows are caused due to illumination variation. An object may have a shadow in a particular illumination but the same object would not have a shadow in a different illumination. This paper deals with removal of shadows using illumination invariant methods based on the fact that an illumination invariant image is a shadow free image. The resultant illumination invariant shadow free image can be used as the input to those applications, where the presence of shadow causes adverse effect. The first step of the method is to find the log chromaticity of image. Next step is to obtain the 1D invariant image. Shadow free chromaticity image is then obtained from the illumination invariant image for better display of the output image. The proposed method is implemented and the results obtained proved the superiority of the method.

More »»

2017

S. Sandheep, John, H., Harikumar, A., and Vinitha Panicker J, “BusTimer: An android based application for generating bus schedules using crowdsourcing”, in 2017 International Conference on Technological Advancements in Power and Energy ( TAP Energy), Kollam, India, 2017.[Abstract]


The fast growth of Indian population has triggered a greater need for well-organized public transport service. Surveys show that approximately 53 million persons travel through almost 400,000 buses twice daily. Often women are seen taking private transportation at night which really compromises their safety. We did some survey and found that most of the women prefer to take public bus transport during night time, but they are forced to take private transport due to lack of an easy means to check the bus schedules. In this paper, we describe and evaluate an android based application that uses crowdsourcing technique for generating bus schedules, even for short distance travel. The system uses the android accelerometer for detecting the bus stops. Then GPS is activated and the coordinates of the stop are sent to the server. The server runs a scheduling algorithm and updates the schedule. And whenever a user of the application enters the source and destination, the corresponding bus schedule will be displayed. The system is implemented and tested using the crowdsourced data populated by twenty five volunteers, as they travelled by public bus transport and the results proved the superiority of the system.

More »»

2015

A. Sadar and Vinitha Panicker J, “DocTool - a tool for visualizing software projects using graph database”, in 2015 Eighth International Conference on Contemporary Computing (IC3), Noida, India , 2015.[Abstract]


In an organization a software development life cycle consists of teams working in different structural hierarchies. Maintaining a complex software where continuous additions and updations are performed by different developers is a challenging task. Also there is a certain amount of latency in communication between two teams regarding an entity of interest in the software. Different software visualization tools have been proposed to address these issues. Many of them provide a view of the software structure by parsing through the source code and analyzing the depth and quality of code. In this paper we propose a DocTool which provide a simple and easy to use solution to two problems: (i) Visualizing the entities of a software and their properties and (ii) Visualizing the workflow in the software. The tool uses a set of json files and a graph database as the backbone. The solution proposed is very simple and provides the user total control over the data he wants to focus on. The tool can be implemented for softwares developed in any kind of platform. The design and implementation of the tool for a Java Web Application software are discussed in this paper.

More »»

2010

Vinitha Panicker J and Wilscy, M., “Detection of moving cast shadows using edge information”, in 2010 The 2nd International Conference on Computer and Automation Engineering (ICCAE), Singapore, Singapore, 2010.[Abstract]


Tracking moving objects in video sequence is an important problem in computer vision, with applications in several fields, such as video surveillance and target tracking. As the shadows attached along with the moving object also have the same motion that of objects, the detection of shadows as foreground objects is very common and produce large errors in object localization and recognition. This paper introduces an effective method which uses the edge information to detect moving cast shadows for traffic sequences. The proposed method initially removes the boundary of the cast shadow, preserving object's interior edges. The coarse object shapes are then reconstructed using the object interior edges. Finally, the cast shadow is detected by subtracting the reconstructed moving object from the change detection mask. The method is implemented and tested using three benchmark videos. The efficiency of the proposed method is compared with five other popular shadow detection methods and the results proved its superiority over others.

More »»