About

• Introduction

  Most of the diseases in plants start from the leaves, and they are the part of the plant where indications of most of the diseases appear first.

  Any digital image needs to be stored in the compressed form, and also decompressed later for further processing

  Due to the popularity and efficiency of the JPEG compression algorithm, JPEG compressed leaf image dataset is used for experimentation

• JPEG compression

  

  Flow Diagram of the JPEG Compression and Decompression steps

  • Compression
  • Decompression

• Proposed Approach

  

  The sample uncompressed disease plant leaves (top) and corresponding images in the DCT compressed domain (bottom)

  The CNN is then trained on both the decompressed images and normal RGB images using the concept of Transfer Learning

  

  A pictorial illustration of the Transfer Learning approach

  

  The flow diagram of the proposed model for plant leaf disease detection in the JPEG compressed domain

• Experimental Results

  • Dataset
  • Results

Info

Title : Detection of Plant Leaf Disease Directly in the JPEG Compressed Domain using Transfer Learning Technique

Authors : Atul Sharma, Bulla Rajesh, Mohammed Javed

Subjects : Computer Vision and Pattern Recognition (cs.CV), Machine Learning (cs.LG), Image and Video Processing (eess.IV)

Submitted on : 10 July 2021

Detection of Plant Leaf Disease Directly in the JPEG Compressed Domain using Transfer Learning Technique