Diagnostics System for Plant Leaf Diseases Using Convergence of Computer Vision and Deep Learning

Diagnostics System for Plant Leaf Diseases Using Convergence of Computer Vision and Deep Learning

Saranya K1 Nasrin begum A2 Archana A3 Kaviya N4 Thamarai selvan M 5

1.   Assistant Professor, Computer Science and Engineering Department, Jai Shriram Engineering College, Tirupur , saranya17113@gmail.com

2.   UG Student, Computer Science and Engineering Department, Jai Shriram Engineering College, Tirupur nasrinbeguma1208@gmail.com

3.  UG Student, Computer Science and Engineering Department, Jai Shriram Engineering College, Tirupur archanacse2020@gmail.com

4.  UG Student, Computer Science and Engineering Department, Jai Shriram Engineering College, Tirupur kaviyan842k03@gmail.com

5.  UG Student, Computer Science and Engineering Department, Jai Shriram Engineering College, Tirupur balaji040203@gmail.com

1. ABSTRACT:
Early and accurate detection of plant diseases is crucial for maximizing agricultural yield and minimizing economic losses.Traditional approaches, however, frequently depend on specialized knowledge.They are ineffective for large-scale applications since they take a lot of time.The suggested method makes use of deep learning models' capabilities.With Densenet-121, it automatically extracts and gains knowledge of pertinent aspects from leaf photos.To identify a bundle of leaves, it employs a multi-object deep learning model.It can decipher and comprehend visual data from the leaves using computer vision.Accuracy is increased by utilizing deep learning and computer vision. The technique of transfer learning allows the system to recognize more veggies than just those in the training dataset. The results of our experimental evaluation show that our method is effective in precisely recognizing vegetables and diagnosing plant leaf diseases. Our system provides a scalable solution for agricultural

management and plant health monitoring by expanding its coverage to include a variety of vegetables and automating disease diagnosis.

KEYWORDS:
CNN,Densenet-121,multi-object model