[4H3-J-13] AI application: agriculture and forestry

[4H3-J-13-04] Study on Appearance Quality Evaluation System of Pear Based on Image Analysis

〇Kosuke Kobayashi¹, tatsuya yamazaki² (1. Niigata University graduate school, 2. Niigata University)

Keywords:Deep Learning, Smart Agriculture, Image Processing, Pear

This paper proposes the system that evaluate an appearance quality of pear based on image analysis. In the proposed method, the appearance deterioration part of pear is detected by using color feature, and deterioration factors are classified by using convolution neural network. In this paper, in order to verify the accuracy of the classification model, experiments using the data set of the appearance deterioration part of pear were performed. As a result of the experiment, the proposed method achieved an average accuracy of 60.50%.