World Journal of Engineering Research and Technology (WJERT) has indexed with various reputed international bodies like : Google Scholar , Index Copernicus , Indian Science Publications , SOCOLAR, China , International Institute of Organized Research (I2OR) , Cosmos Impact Factor , Research Bible, Fuchu, Tokyo. JAPAN , Scientific Indexing Services (SIS) , Jour Informatics (Under Process) , UDLedge Science Citation Index , International Impact Factor Services , International Scientific Indexing, UAE , International Society for Research Activity (ISRA) Journal Impact Factor (JIF) , International Innovative Journal Impact Factor (IIJIF) , Science Library Index, Dubai, United Arab Emirates , Scientific Journal Impact Factor (SJIF) , Science Library Index, Dubai, United Arab Emirates , Eurasian Scientific Journal Index (ESJI) , Global Impact Factor (0.342) , IFSIJ Measure of Journal Quality , Web of Science Group (Under Process) , Directory of Research Journals Indexing , Scholar Article Journal Index (SAJI) , International Scientific Indexing ( ISI ) , Scope Database , Academia , 

World Journal of Engineering Research and Technology

( An ISO 9001:2015 Certified International Journal )

An International Peer Reviewed Journal for Engineering Research and Technology

An Official Publication of Society for Advance Healthcare Research (Reg. No. : 01/01/01/31674/16)

ISSN 2454-695X

Impact Factor : 7.029

ICV : 79.45

News & Updation

  • Article Invited for Publication

    Article are invited for publication in WJERT Coming Issue

  • ICV

    WJERT Rank with Index Copernicus Value 79.45 due to high reputation at International Level

  • New Issue Published

    Its Our pleasure to inform you that, WJERT March 2024 Issue has been Published, Kindly check it on https://www.wjert.org/home/current_issues

  • WJERT New Impact Factor

    WJERT Impact Factor has been Increased to 7.029 for Year 2024.

  • WJERT: MARCH ISSUE PUBLISHED

    March 2024 Issue has been successfully launched on 1 March 2024.

Indexing

Abstract

A STUDY ON THE METHODS OF IMAGE FEATURE EXTRACTION

*Sayeed Mohammed Shaik and Dr. R. Kamalraj

ABSTRACT

Computer Vision tasks analyze the whole image. These tasks are computationally expensive. By extracting features from the image, we can reduce the computational expense. Using Deep Learning we can create Intelligent Models that can perform these Computer Vision tasks. We can tackle each task such as image classification and object detection with Models specifically trained to deal with those tasks. By combining said specifically trained Models we can achieve an aggregated output. Some of the techniques such as Instance Segmentation deal with the Computer Vision tasks in similar fashion. Features in deep learning can range from luminance of the pixels to high level pattern classification. Images are represented as an array of pixels, we can utilize Convolutional Neural Networks for the Image Processing removing the need to transform the array of pixels. Other Neural Network architectures such as U-net and Auto-Encoders can also be utilized in dealing with similar Computer Vision tasks. Many Models are trained to deal with specific tasks. This paper surveys some of the Models currently performing the Computer Vision tasks and Image processing.

[Full Text Article] [Download Certificate]