• MANEM SRAVANI
• SARAGADAM SRIDHAR

In the agricultural industry, the detection of rotten fruits holds great significance. While humans can classify fresh and rotten fruits, their effectiveness diminishes due to fatigue from repetitive tasks. On the other hand, machines do not tire, making them ideal for such tasks. Fruits play a crucial role in maintaining a healthy lifestyle, as they provide essential nutrients and vitamins. However, the quality of fruits can vary, with some being fresh and ripe while others may be rotten or spoiled. The objective of this project is to employ machine learning techniques to accurately classify fresh and rotten fruits. To address this issue, a proposed solution aims to minimize human effort, production time, and cost by identifying fruit defects. The failure to identify defects could result in the contamination of good fruits, highlighting the necessity for a model to prevent the spread of fruit diseases. The proposed model utilizes a Convolutional Neural Network (CNN) to extract relevant features from fruit images. These features are then processed using the Softmax algorithm for classification into fresh or rotten categories. The model's performance was evaluated using a dataset obtained from Kaggle and demonstrated an impressive accuracy of 96% on the test dataset. The results clearly indicate that the CNN model is highly effective in accurately classifying fresh and rotten fruits. Furthermore, transfer learning methods were explored and compared, revealing that the proposed CNN model outperformed both transfer learning and state-of-the-art models in terms of classification accuracy. Through the utilization of a mobile app equipped with deep learning techniques, farmers can now detect fruit diseases with great precision. This app not only performs fruit disease classification but also incorporates a portable and highly accurate deep learning model. In conclusion, this project successfully demonstrates the efficacy of using machine learning, particularly a CNN model, in accurately classifying fresh and rotten fruits. The development of a mobile app utilizing deep learning techniques provides farmers with a convenient and reliable tool for detecting fruit diseases and ensuring the overall quality of their produce.

CNN (Convolutional Neural Network)

"DISEASED FRUIT CLASSIFICATION USING MACHINE LEARNING AND PYTHON", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.10, Issue 7, page no.d448-d514, July-2023, Available :http://www.jetir.org/papers/JETIRTHE2053.pdf

"DISEASED FRUIT CLASSIFICATION USING MACHINE LEARNING AND PYTHON", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.10, Issue 7, page no. ppd448-d514, July-2023, Available at : http://www.jetir.org/papers/JETIRTHE2053.pdf