A flight sensing detector for apple maturity indexes in orchard

Apples quality monitoring and maturity measurement are important for yield evaluation, profit estimation, determination of harvest time and postharvest storage conditions. This study explored the monitoring ability of the micro-Unmanned Aerial Vehicle (UAV) to the orchard, the force sensor and gyroscope are used to observe the force and posture changes of apples during the spectral collection process. The UAV carries 18 channels of visible/ near-infrared multispectral sensor to collect the internal information of growing apples at close range. A total of 200 apples were collected in the orchard, and then soluble solid content (SSC), titratable acid (TA), firmness, and starch pattern index (SPI) were destruc-tively tested. Multiple scatter correction (MSC) and Standard normal variate (SNV) pre-treatment combined with partial least squares regression (PLSR) were used to establish the internal quality prediction model. Streif index is used to evaluate the maturity of apples. The results showed that the detection of apples on trees by UAV would not affect the growth of apples and could collect the internal quality information. The correlation coef-ficient of the prediction model for apple maturity is 0.8593 and root mean square error is 0.0102. It shows that this study has potential in detecting the changes of quality and maturity of growing apples and is helpful to the overall perception of orchards.& COPY; 2023 IAgrE. Published by Elsevier Ltd. All rights reserved.

Automated summary

This study explored the monitoring ability of a micro-Unmanned Aerial Vehicle (UAV) in orchards by using a force sensor and gyroscope to observe changes in force and posture of growing apples during spectral collection. The UAV carried a visible/near-infrared multispectral sensor to collect internal information of the apples. A prediction model for internal quality was established using destructive testing and pre-treatments, and apple maturity was evaluated using the Streif index. Results showed that detecting apples on trees using UAV did not affect their growth and provided valuable internal quality information. The prediction model had a correlation coefficient of 0.8593 and a root mean square error of 0.0102, indicating the potential of this study in detecting quality and maturity changes in growing apples and its usefulness in overall orchard perception.