Technical Name Development of fruit tree industry monitoring technology based on multi-source image recognition technology
Project Operator Taiwan Agricultural Research Institute Council of Agriculture, Executive Yuan
Project Host 郭鴻裕, 陳淑佩, 林毓雯, 周呈霙, 王驥魁
Summary
Integrating deep learning, 3D information analysis, hyperspectral analysis, computer vision analysiscombined the multi-source images to develop fruit quality monitoring techniques, including: planting area monitoring, position monitoring, yield monitoring, harvesting time prediction, fruit maturity measurement, fruit quality testing, to achieve the goal of enhancing industrial value.
Scientific Breakthrough
Identification system applied to fruit trees classification can reach mean average precision of 0.85. Airborne LiDAR can penetrate the screen-house to obtain the fruit trees information inside.
Monitoring the production seasonrelease yield prediction monthly.
Automatic fruit selection reduces human’s screening,we use hyperspectral technology to measure fruit maturity.
Industrial Applicability
AgricultureFood Agency:  Interpret fruit treescalculate area.
BAPHIQ:  Apply height data for Tessaratoma papillosa Drury.
Agribusiness:  Farm monitoring.
Council of Agriculture:  Keep yieldprice in balance by yield prediction.
Industry:  We cooperated with Farmer’s Associationtwo companies for automatic fruit selectionautomatic interpretation of fruit maturity.
Keyword Fruit Tree Identification Image Segmentation Machine Learning Convolutional Neural Network Airborne LiDAR Hyperspectral Analysis Yield Predict Quality Inspection Land Usage Agricultural Monitering