i-manager's Journal on Image Processing (JIP)


Volume 8 Issue 3 July - September 2021

Research Paper

Identifying Drowning Objects in Flood Water and Classifying using Deep Convolution Neural Networks

Kovvuri N. Bhargavi* , G. Jaya Suma**
*Department of Information Technology, Aditya College of Engineering and Technology, Surampalem, Andhra Pradesh, India.
** Department of Information Technology, JNTUK-UCEV, Vizianagaram, Andhra Pradesh, India.
Bhargavi, K. N., and Suma, G. J. (2021). Identifying Drowning Objects in Flood Water and Classifying using Deep Convolution Neural Networks. i-manager's Journal on Image Processing, 8(3), 1-14. https://doi.org/10.26634/jip.8.3.18451

Abstract

Natural disasters like floods often occur due to heavy rainfall, storms, melting snow and ice, overflowing rivers, dam failures, and urban drainage systems. Failing to evacuate flooded watery areas leads to drowning of objects into the water. While drowning in water people feel difficult to breathe and may not survive for a long period. Real-time Detection of persons and vehicles in heavy water flow is a challenging task. This paper proposes a framework to identify floating and almost drowned objects in the water and classify them whether they are humans or non-living objects using a series of Convolution Neural Network object detection models. Faster RCNN, Mask RCNN and, You Only Look Once (YOLOv5) network models are trained on the image dataset. In case of overlapped objects, object segmentation is also performed using Mask RCNN for predicting the shape of the drowning object. Faster RCNN and YOLOv5 models are validated using a test dataset and a decline in training loss is plotted on a tensor board. Results for evaluating object detection model show Faster RCNN as the best model for detecting and classifying objects in water than YOLOv5 and Mask RCNN. Distance between object is measured to find the shortest path to reach the object for faster rescue operations. Counting and Tracking of objects are performed to know the exact count of objects who need help in an emergency and to monitor their real-time position in the water.

Research Paper

Bank Transaction using Iris Recognition System

Nithesh Kumar R.* , Rahul D.**, V. Dhanakoti***, Saran S.****
*-****Department of Computer Science and Engineering, SRM Valliammai Engineering College, Kanchipuram, Tamil Nadu, India.
Kumar, R. N., Rahul, D., Dhanakoti, V., and Saran, S. (2021). Bank Transaction using Iris Recognition System. i-manager's Journal on Image Processing, 8(3), 15-20. https://doi.org/10.26634/jip.8.3.18124

Abstract

This paper targets to develop a powerful algorithm for transacting money with high level security and high recognition rates in various environments. Haar Cascade algorithm has been applied for fast and easy face detection from the input image. The face image is then being converted into grayscale image. Finally the iris of the users are paired up and also the cost of each possible pairing is computed by a mixture of mathematical models. With human interaction, voice commands are used for transactions.

Research Paper

Implementation of Approximate Adders and Multipliers for Error Tolerant Image Processing

P. Mani* , S. Priyadharshini**, N. Priyanga***, V. Reshma****
*-****Department of Electronics and Communication Engineering, K. Ramakrishnan College of Technology, Trichy, Tamil Nadu, India.
Mani, P., Priyadharshini, S., Priyanga, N., and Reshma, V. (2021). Implementation of Approximate Adders and Multipliers for Error Tolerant Image Processing. i-manager's Journal on Image Processing, 8(3), 21-26. https://doi.org/10.26634/jip.8.3.18218

Abstract

An adder is the basic computational circuit in Very Large Scale Integration (VLSI) digital design. To improve the design metrics of an adder, Approximate Adders (AAs) have been proposed. These adders have been applied and analyzed on 8x8 Dadda Multipliers (DMs). The design metrics of proposed AAs, Approximate Dadda Multipliers (ADMs) are synthesized in Cadence Register-Transfer Level (RTL) compiler and compares the design metrics with three different technology nodes. The HDL synthesis results shows that the delay to the output of the proposed 8-tap filter gains an improvement over the conventional method. The implementation is done using Verilog HDL. Simulation and synthesis are done with the Xilinx ISE tool.

Research Paper

Lane Detection System to Avoid Accidents using Image Processing and Deep Learning

Shivani* , Vishnu Priya**, Yogapriya***, Bhagyalakshmi****, Vanshika*****
*-*****Department of Electronics and Communication Engineering, Rajalakshmi Engineering College, Chennai, Tamil Nadu, India.
Shivani, Priya, V., Yogapriya, Bhagyalakshmi, and Vanshika. (2021). Lane Detection System to Avoid Accidents using Image Processing and Deep Learning. i-manager's Journal on Image Processing, 8(3), 27-31. https://doi.org/10.26634/jip.8.3.18193

Abstract

To reduce the frequency of road accidents, Lane detection system can be implemented in four wheelers to identify lane borders on the road and further prompt the driver if he switches and moves to erroneous lane markings. There are many algorithms used in this system namely, perspective, sliding window algorithms and many more, so collectively these can be called as a library. Lane boundaries are detected using a camera that captures the view of the road. The camera mounted on the top of the car takes images which serves as the basic input.

Research Paper

Labview for Motion Detection using Webcam

Sushma Thotakura* , Baby Harshithakanneganti**
*-**Department of Electronics and Communication Engineering, Prasad V. Potluri Siddhartha Institute of Technology, Vijayawada, Andhra Pradesh, India.
Thotakura, S., and Harshithakanneganti, B. (2021). Labview for Motion Detection using Webcam. i-manager's Journal on Image Processing, 8(3), 32-36. https://doi.org/10.26634/jip.8.3.18235

Abstract

In today’s world, security plays a major role in day-to-day life. In general, sensors are used to detect an intruder in secure places. The main problem with the sensor arises because of its specifications, limitations in aspect of temperature and humidity etc., In this paper, webcam is used instead of sensor to detect the moving object using myRIO. The motion of objects is detected by computing the difference between the images captured by the webcam. The detection process is achieved using LabVIEW software and myRIO