Munish Kumar

Maharaja Ranjit Singh Punjab Technical University

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Trends

Author Order

Document Type

Co-Authors

M. K. Jindal

Panjab University

Krishan Kumar

Guru Jambheshwar University of Science and Technology

Harmandeep Kaur

Guru Nanak Dev University

Simpel Rani Jindal

Punjabi University

Shaveta Dargan

Maharaja Ranjit Singh Punjab Technical University

Chella Santhosh

Koneru Lakshmaiah Education Foundation

Rajendra Kumar Sharma

Thapar Institute of Engineering & Technology

Rupinder Kaur

Punjabi University

Naresh Kumar Garg

Indian Institute of Technology Kanpur

Ajay Mittal

Panjab University

Cooperative Institutions

Koneru Lakshmaiah Education Foundation

Chandigarh University

Thapar Institute of Engineering & Technology

Panjab University

Maharaja Ranjit Singh Punjab Technical University

Punjabi University

Chitkara University

Bhabha Atomic Research Centre

Institute of Engineering

Lovely Professional University

Author Statistics

Papers

Citation

H-Index

i-10 index

Research Field

Size of Training Set Vis-à-vis Recognition Accuracy of Handwritten Character Recognition System

Journal of Emerging Technologies in Web Intelligence (2013)

Munish Kumar Richa Sharma M. K. Jindal

Support Vector Machines (SVMs) have successfully been used for character recognition. In the present study, we have shown how the recognition accuracy of a SVM classifier varies with variation in the training set size. The training set for this work is taken from samples of offline handwritten Gurmukhi characters. For recognition of a handwritten Gurmukhi character, we have used curvature features extracted from the skeletonized image of each Gurmukhi character. Features of a character have been computed based on statistical measures of distribution of points on the bitmap image of character. To extract these features, the image of each Gurmukhi character is first segmented into few zones and then the curvature shape is computed within each of these zones. Considering all the zones, a feature set is formed for representation of each image pattern and a database of 3500 isolated handwritten Gurmukhi characters has been used for the same. The results of investigation presented in this paper show that the size of training set has a significant effect on the accuracy of offline handwritten Gurmukhi script recognition system. Index Terms—Feature extraction; curve fitting; handwritten character recognition; SVM.

Training set

Character encoding

10.4304/jetwi.5.4.380-384

Cite

Citations (3)

MDP Feature Extraction Technique for Offline Handwritten Gurmukhi Character Recognition

The Smart Computing Review (2013)

Munish Kumar

Character recognition is intricate work because of the various writing styles of different individuals. Most of the published work on handwritten character recognition problems deals with statistical features, and a few works deal with structural features, in general, and Gurmukhi script, in particular. In the present work, we propose a methodology for offline handwritten Gurmukhi character recognition by using a modified division points (MDP) feature extraction technique. We also compare this technique with other recently used feature extraction techniques, namely zoning features, diagonal features, directional features, intersection and open end points features, and transition features. To select a representative set of features is the most significant task for a character recognition system. After feature extraction, the classification stage makes use of the features extracted in the previous stage to recognize the character. In this work, we used linear-support vector machines (linear-SVM), k-nearest neighbor (k-NN), and multilayer perceptron (MLP) classifiers for recognition. For experimental analysis, we used 10,500 samples of the isolated, offline, handwritten, basic 35 akhars of Gurmukhi script. The proposed system achieved a maximum recognition accuracy of 84.57%, 85.85% and 89.20% with linear-SVM, MLP and k-NN classifiers, respectively, with a five-fold cross validation technique.

Feature (linguistics)

10.6029/smartcr.2013.06.001

Cite

Citations (16)

High-Quality in Data Authentication Dodging Massive Attack in VANETS

International Journal on Recent and Innovation Trends in Computing and Communication (2019)

Munish Kumar Maninder Kaur Inderdeep Kaur

VANET plays an important role in the Security terms. VANET network is due to their unique features like as a high dynamic network (topology) and Mobility prediction. It attracts so much attention to the industry. VANET wireless networks are rapidly increased commercial and academic interests. Mobile connectivity, Traffic congestion management and road safety are some applications that have arisen within this network model. The routing protocol is a reactive type which means if there is data to be sent then the way will create. On-demand Distance Vector routing protocol is a generally used network topology based on rules for VANET. In surveyed of the routing protocol implemented a balance AODV method used for identifying the malicious nodes in the network. A balanced AODV routing method is defined with following characteristics:- (i) Use of threshold adaptive according to the network situations and balance index i.e node nature. (ii) Detect the malicious node in the network. (iii) Detection and prevention methods in real-time and independent on each vehicle node. In research paper, implement a B-AODV routing protocol and RSA method for detection and prevention the malicious node in the vehicular network. In this proposed algorithm, each vehicle node is employing balance index for acceptable and reject able REQ information’s (Bits). The consequences of the simulation tool in MATLAB (Matrix Laboratory) indicates BAODV and RSA method is used to detect and prevent the flood attach and loss of network bandwidth. Comparison between AODV, BAODV, RSA in normal phase defines B-AODV is exactly matched with AODV in the vehicular network and performance analysis overhead, an end to end delay and packet delivery rate.

Distance-vector routing protocol

10.17762/ijritcc.v7i5.5303

Cite

Citations (1)

Improved Multi-Face Detection with ResNet for Real-World Applications

Munish Kumar Monali Gulhane Sandeep Kumar Harshita Sharma Rashmi Verma

This paper presents an improved approach to multi-face detection using ResNet50 in Python for real-world applications. The study focuses on enhancing the accuracy and efficiency of detecting multiple faces in a single frame, a challenge often encountered in real-world scenarios such as surveillance and social media platforms. The proposed method leverages the power of ResNet, a deep residual learning framework known for its superior performance in image recognition tasks. The paper discusses the implementation details of the ResNet model in Python and how it has been optimized for multi-face detection. Experimental results demonstrate that the improved ResNet model outperforms traditional face detection methods regarding accuracy and speed, making it a promising solution for real-world applications. The paper concludes with potential future work and improvements that could further enhance the model's performance.

Residual neural network

10.1109/smart59791.2023.10428518

Cite

Citations (0)

A Novel Technique to Segment Maps from Scanned Document Images

Advances in Science, Technology & Innovation/Advances in science, technology & innovation (2025)

Rupinderpal Kaur Munish Kumar Manish Kumar Jindal

10.1007/978-3-031-72747-4_33

Cite

Citations (0)

Semi-supervised labeling: a proposed methodology for labeling the twitter datasets

Multimedia Tools and Applications (2022)

Tabassum Gull Jan Surinder Singh Khurana Munish Kumar

Microblogging

Supervised Learning

Labeled data

Sentiment Analysis

Majority Rule

10.1007/s11042-022-12221-7

Cite

Citations (3)

DeepNet-Gait: Human Identification by Gait Using Convolutional Neural Network Model

Veenu Rani Munish Kumar

In the last decade, human identification has been considered as one of the most important aspects of social security.Biometrics, Which come under the shadow of pattern recognition,are used to identify a human by matching various patterns in an image or video. For example, fingerprint identification is an imagepattern recognition application that is widely used to authenticate a person while working with smartphones and for attendance marking. In a similar way, a gait recognition system is a video pattern identification application that is used to recognize a personfrom a distance. In this article, the authors used a deep learning CNN model to extract as well as classify features. The CASIA-A dataset was used in the experiments. This dataset contains images of 20 people. Each person has 12 images. Our suggested model acquired an overall accuracy of 99.28%, suggesting that it is the most accurate technique when compared to other state-of-theart studies.

Identification

10.1109/spin57001.2023.10117067

Cite

Citations (3)

FEMT: a computational approach for fog elimination using multiple thresholds

Multimedia Tools and Applications (2020)

Mamta Mittal Munish Kumar Amit Verma Iqbaldeep Kaur Bhavneet Kaur

Haze

Visibility

Ground truth

10.1007/s11042-020-09657-0

Cite

Citations (7)

DeepNetDevanagari: a deep learning model for Devanagari ancient character recognition

Multimedia Tools and Applications (2021)

Sonika Rani Narang Munish Kumar M. K. Jindal

Devanagari

Extractor

Feature (linguistics)

Optical character recognition

10.1007/s11042-021-10775-6

Cite

Citations (65)

Image Reconstruction Using Impedance Measurement as EIT Technique

IOSR Journal of Electrical and Electronics Engineering (2013)

Munish Kumar

Electrical Impedance Tomography (EIT) is a radiation-less imaging technology and offers a noninvasive imaging system for human organs.EIT is fast, inexpensive and portable as compared to other imaging techniques.The presented work is based on potential profile computed from the raw data as obtained from the body surface via the electrodes array placed around the organ.A 16 electrode array is proposed for extracting the potential profile.The potential profile is obtained in digital format by sing a 24 bit ADC in order to get a higher resolution potential profile.This helps in reconstructing the organ image at highly resolution.

Electrical Impedance Tomography

10.9790/1676-0718487

Cite

Citations (0)