Artificial Intelligence, Computer Vision and Pattern Recognition
5
Scopus Publications
48
Scholar Citations
1
Scholar h-index
1
Scholar i10-index
Scopus Publications
Accurate Measurement of Nail Overgrowth with Twin YOLOv8 Segmentation Models for Gamified Disease Prevention in Rural Schools Aravindhan R, Pradeep Kumar R, Sunanth H, Arun Karthik V, Anusha T Proceedings of the IEEE International Conference on AI Engineering and Innovations Aiei 2026, 2026 A positive hygiene habit in children is essential to prevent the transmission of pathogens but it remains a challenge. Middle school children in rural areas are prone to have poor hand hygiene due to lack of knowledge, independent care practices and attitude. Studies suggest entertainment-based educational intervention helps to improve awareness and reduces the risk of illness. This paper proposes a gamified web application, Smart-Groom, to analyze fingernail overgrowth and educate children in the most effective manner. This paper proposes a novel Dual Segmentation Caliper (DSC) algorithm to quantify nail overgrowth from a single color image. Previous works rely on direct edge detection or skin-contour analysis but they fail as the nail tip occupies less than 1% of the image. DSC utilizes two distinct YOLOv8-seg models that runs in parallel, where one is trained to segment the whole nail unit and the other to segment the nail plate (pink bed). By anchoring these detections within the defined Regions of Interest (ROIs), we calculate overgrowth through differential coordinate analysis. Experimental results demonstrate that this twin-network approach achieves accurate nail overgrowth estimation with a mAP50 of 0.993, enabling submillimetre measurement accuracy (MAE ± 0.3 mm).
Efficient Communication Model for a Smart Parking System with Multiple Data Consumers T. Anusha, M. Pushpalatha Smart Cities, 2022 A smart parking system (SPS) is an integral part of smart cities where Internet of Things (IoT) technology provides many innovative urban digital solutions. It offers hassle-free parking convenience to the city dwellers, metering facilities, and a revenue source for businesses, and it also protects the environment by cutting down drive-around emissions. The real-time availability information of parking slots and the duration of occupancy are valuable data utilized by multiple sectors such as parking management, charging electric vehicles (EV), car servicing, urban infrastructure planning, traffic regulation, etc. IPv6 wireless mesh networks are a good choice to implement a fail-safe, low-power and Internet protocol (IP)-based secure communication infrastructure for connecting heterogeneous IoT devices. In a smart parking lot, there could be a variety of local IoT devices that consume the occupancy data generated from the parking sensors. For instance, there could be a central parking management system, ticketing booths, display boards showing a count of free slots and color-coded lights indicating visual clues for vacancy. Apart from this, there are remote user applications that access occupancy data from browsers and mobile phones over the Internet. Both the types of data consumers need not collect their inputs from the cloud, as it is beneficial to offer local data within the network. Hence, an SPS with multiple data consumers needs an efficient communication model that provides reliable data transfers among producers and consumers while minimizing the overall energy consumption and data transit time. This paper explores different SPS communication models by varying the number of occupancy data collators, their positions, hybrid power cycles and data aggregation strategies. In addition, it proposes a concise data format for effective data dissemination. Based on the simulation studies, a multi-collator model along with a data superimposition technique is found to be the best for realizing an efficient smart parking system.
A machine learning approach in communication 5G-6G network SL Boppana, A Gummadi, YR Raju, S Thatavarti, R Kuchipudi, ... J. Theor. Appl. Inf. Technol 102 (10), 4268 , 2024 2024.0 Citations: 1
Blockchain technology for fraudulent practices in insurance claim process J Gera, AR Palakayala, VKK Rejeti, T Anusha 2020 5th international conference on communication and electronics systems … , 2020 2020.0 Citations: 45
A New Digital Encryption Scheme: Binary Matrix Rotations Encryption Algorithm KS Kumar, M Srikanth, T Anusha International Journal of Research 2 (2), 18-27 , 2015 2015.0 Citations: 1
Handwritten Isolated Digit Recognition Using Artificial Neural Networks KS Kumar, T Anusha, B Ramesh International Journal Of Computer Science Applications & Information … , 2013 2013.0 Citations: 1
PESTICIDES SPRAY DRONE DESIGN FOR SMART AGRICULTURE RY Rao, Y Raju, T Anusha, T Sreeker, T Hema, E Blessy
MOST CITED SCHOLAR PUBLICATIONS
Blockchain technology for fraudulent practices in insurance claim process J Gera, AR Palakayala, VKK Rejeti, T Anusha 2020 5th international conference on communication and electronics systems … , 2020 2020.0 Citations: 45
A machine learning approach in communication 5G-6G network SL Boppana, A Gummadi, YR Raju, S Thatavarti, R Kuchipudi, ... J. Theor. Appl. Inf. Technol 102 (10), 4268 , 2024 2024.0 Citations: 1
A New Digital Encryption Scheme: Binary Matrix Rotations Encryption Algorithm KS Kumar, M Srikanth, T Anusha International Journal of Research 2 (2), 18-27 , 2015 2015.0 Citations: 1
Handwritten Isolated Digit Recognition Using Artificial Neural Networks KS Kumar, T Anusha, B Ramesh International Journal Of Computer Science Applications & Information … , 2013 2013.0 Citations: 1
PESTICIDES SPRAY DRONE DESIGN FOR SMART AGRICULTURE RY Rao, Y Raju, T Anusha, T Sreeker, T Hema, E Blessy
