Sensor-based in-situ detector for distinguishing between forchlorfenuron treated and untreated kiwifruit at multi-wavelengths
Sensor-based in-situ detector for distinguishing between forchlorfenuron treated and untreated kiwifruit at multi-wavelengths
复制标题
基于传感器的原位检测器,用于在多波长下区分经氯吡脲处理和未经处理的猕猴桃
DOI:
10.1016/j.biosystemseng.2019.11.019
复制
发表时间:
2020-02-01
影响因子:
5.1
通讯作者:
Zhu, Zhuozhuo
中科院分区:
文献类型:
--
作者:
Guo, Wenchuan;Wang, Kai;Zhu, Zhuozhuo
Forchlorfenuron (CPPU) can increase the production of kiwifruit, but it decreases the fruit quality and shelf life, and increases the rotting rate during storage. Therefore, identifying the kiwifruit treated with CPPU is an urgent task to be solved in the kiwifruit industry. This work presents a convenient, low-cost, and sensor-based handheld detector for identifying kiwifruit treated with CPPU by measuring reflectance at five wavelengths (660, 940, 1064, 1250 and 1445 nm), which were selected from 931 nm to 1689 nm and determined by available optical filters on the market. The detector consisted of a micro-controller, a light source, a multi-wavelength selection apparatus, a light detection and signal processing module, a display and a battery. The multi-wavelength selection apparatus was carried out by five optical filters, and the reflected light was sensed by five photodiodes. The whole system was controlled using a STM32 micro-controller. Partial least squares discrimination analysis was used to build the model for distinguishing kiwifruit between CPPU treated and untreated. The detector was tested on two varieties of kiwifruit ('Huayou' and 'Xuxiang') treated with different concentration levels of CPPU. The test showed that the detector could offer accuracies of 92.9% and 85.8% for 'Huayou' and 'Xuxiang', respectively, and the detection result can be given in 2 s. The study makes identifying kiwifruit treated with CPPU be possible, and it offers valuable information for developing detectors for other fruits treated with CPPU, such as grape, strawberry, etc. (C) 2019 IAgrE. Published by Elsevier Ltd. All rights reserved.