Please use this identifier to cite or link to this item: http://hdl.handle.net/10603/479609
Title: Design of efficient and effective deep Learning architectures for processing the agricultural images to improve the performance of smart farming
Researcher: Arumuga Arun, R
Guide(s): Umamaheswari, S
Keywords: Engineering and Technology
Computer Science
Computer Science Information Systems
Design of efficient and effective
deep Learning architectures
improve the performance of smart farming
University: Anna University
Completed Date: 2023
Abstract: Agriculture, a powerful word that requires no definition, contributes the most profitable share of the Indian Economy. On a wider view, intensifying productivity requires very basic things like improving the quality and quantity of the crop yield and dwindling the agricultural expenses. There are certain cases that affect productivity, among which the presence of weeds, crop diseases, and pests in crop fields stands top. Traditional methods, such as manual removal of weeds and spraying of agrochemical products like herbicides, and pesticides have chances of harming the crops and also add to the expenses. Selective treatment against weeds, crop diseases, and pests is a cost-effective method that reduces manpower and usage of the agrochemical, at the same time it requires an effective computer vision system to identify issues and should be smaller in size to run in the resource-constrained device. Machine learning-based activities for identifying weed portions, crop-disease classification, and pest detections are tedious and time-consuming processes. Even though existing deep learning-based approaches perform well, they are hard to deploy on resource-constrained devices commonly used by farmers due to their high computational cost. newlineThis research work presents four contributions, where the first three contributions to handling the challenges of facilitating selective treatment against weeds, crop diseases, and pests, and the fourth contribution is the development of a lightweight computer vision-based Deep Learning based Smart Farming Application. The first contribution presents a method of effective crop weed segmentation in the agricultural field. newline
Pagination: xxiii,174p.
URI: http://hdl.handle.net/10603/479609
Appears in Departments:Faculty of Information and Communication Engineering

Files in This Item:
File Description SizeFormat 
01_title.pdfAttached File102.58 kBAdobe PDFView/Open
02_prelim pages.pdf638.64 kBAdobe PDFView/Open
03_content.pdf448.03 kBAdobe PDFView/Open
04_abstract.pdf340.04 kBAdobe PDFView/Open
05_chapter 1.pdf1.2 MBAdobe PDFView/Open
06_chapter 2.pdf678.67 kBAdobe PDFView/Open
07_chapter 3.pdf1.23 MBAdobe PDFView/Open
08_chapter 4.pdf1.25 MBAdobe PDFView/Open
09_chapter 5.pdf1.46 MBAdobe PDFView/Open
10_chapter 6.pdf1.06 MBAdobe PDFView/Open
11_anneuxres.pdf277.86 kBAdobe PDFView/Open
80_recommendation.pdf172.22 kBAdobe PDFView/Open
Show full item record


Items in Shodhganga are licensed under Creative Commons Licence Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0).

Altmetric Badge: