Computer vision is concerned with the automatic extraction, analysis, and understanding of useful information from a single image or a sequence of images. We have used Convolutional Neural Networks (CNN) in automatic image classification systems. In most cases, we utilize the features from the top layer of the CNN for classification; however, those features may not contain enough useful information to predict an image correctly. In some cases, features from the lower layer carry more discriminative power than those from the top. Therefore, applying features from a specific layer only to classification seems to be a process that does not utilize learned CNN’s potential discriminant power to its full extent. Because of this property we are in need of fusion of features from multiple layers. We want to create a model with multiple layers that will be able to recognize and classify the images. We want to complete our model by using the concepts of Convolutional Neural Network and CIFAR-10 dataset. Moreover, we will show how MatConvNet can be used to implement our model with CPU training as well as less training time. The objective of our work is to learn and practically apply the concepts of Convolutional Neural Network.
18 May 2019 in Global Journal of Computer Science and Technology
Humans can see and visually sense the world around them by using their eyes and brains. Computer vision works on enabling computers to see and process images in the same way that human vision does. Several algorithms developed in the area of computer vision to recognize images. The goal of our work will be to create a model that will be able to identify and determine the handwritten digit from its image with better accuracy. We aim to complete this by using the concepts of Convolutional Neural Network and MNIST dataset. We will also show how MatConvNet can be used to implement our model with CPU training as well as less training time. Though the goal is to create a model which can recognize the digits, we can extend it for letters and then a person’s handwriting. Through this work, we aim to learn and practically apply the concepts of Convolutional Neural Networks.
21 April 2025 in Journal of Agriculture and Food Research
An attacker can attack a network in several methods when there are a lot of device connections. Distributed denial-of-service (DDoS) attacks could result from this circumstance, which could damage resources and corrupt data. Therefore, irregularity in traffic data must be detected to identify malicious behavior in a network, which is critical for maintaining the integrity of current cyber-physical systems (CPS) as well as network security. This article attempts to study and compare various approaches to detecting DDoS attacks and expresses data paths for packet filtering for high-speed networks (HSN) performance, using machine or deep learning techniques used in intrusion detection systems (IDSs) and flow-based IDSs. The study presents a comprehensive DDoS attack taxonomy, categorizes detection strategies, and highlights the HSN accuracy assessment features. By exposing the problems and difficulties associated with DDoS attacks on HSN, several investigation paths are proposed to assist researchers in determining and developing the best solution.
21 May 2019 in Global Journal of Computer Science and Technology
This paper presents an innovative design for Optical Character Recognition (OCR) from text images by using the Template Matching method.OCR is an important research area and one of the most successful applications of technology in the field of pattern recognition and artificial intelligence.OCR provides full alphanumeric visualization of printed and handwritten characters by scanning text images and converts it into a corresponding editable text document. The main objective of this system prototype is to develop a prototype for the OCR system and to implement The Template Matching algorithm for provoking the system prototype. In this paper, we took alphabet (A-Z and a-z), and numbers (0-1), grayscale images, bitmap image format were used and recognized the alphabet and numbers by comparing between two images. Besides, we checked accuracy for different fonts of alphabet and numbers. Here we used Matlab R 2018 a software for the proper implementation of the system.
8 August 2024 in International Journal of Imminent Science & Technology.
In recent years, the development of cross-platform applications has gained increasing popularity due to their ability to enhance accessibility, usability, and functionality across multiple devices. This paper proposes a cross-platform vehicle tracking system for Pabna University of Science and Technology (PUST), addressing key challenges in vehicle management, transportation efficiency, and security. The system includes both android and web interfaces designed to provide real-time location monitoring of university vehicles. The primary problems identified are the inefficient tracking of university vehicles, lack of security in vehicle usage, and suboptimal route management. To address these issues, our system leverages modern technologies such as the Google Maps API, Node.js, and the Real-time Firebase Database. The Android app and web-based interfaces allow authorized users to track the real-time location of university vehicles, access detailed vehicle information, and review historical routes. Additionally, the system issues real-time alerts for unauthorized usage or irregular activities. The embedded system uses a Global Positioning System (GPS) and Global System for Mobile Communication (GSM) for tracking and positioning vehicles. The ESP8266 NodeMCU is interfaced serially with a GPS module Ublox NEO6M Receiver to continuously monitor and report vehicle status. The GPS module sends the vehicle’s position (latitude and longitude) to the real-time Firebase database, which then transmits this data to the mobile and web interfaces. The proposed system underwent comprehensive user acceptance tests, yielding satisfactory results that attest to its potential for enhancing vehicle management, transportation efficiency, and security at PUST. This paper details the design, development, and evaluation of the system, highlighting its benefits and outlining future research directions.