RESUMEN
Moisture content testing of agricultural products is critical for quality control, processing efficiency and storage management. Testing foxtail millet moisture content ensures stable foxtail millet quality and helps farmers determine the best time to harvest. A differential capacitance moisture content detection device was designed based on STM32 and PCAP01 capacitance digital converter chip. The capacitance method combined with the back-propagation(BP) algorithm and the extreme learning machine(ELM) algorithm was chosen to construct an analytical model for foxtail millet moisture content, temperature, and volume duty cycle. This work performs capacitance measurements on foxtail millet with different moisture contents, temperatures, and proportions of the measured substance occupying the detection area (that is, the volumetric duty cycle). On this foundation, the sparrow search algorithm (SSA) is used to optimize the BP and ELM models. However, SSA may encounter problems such as falling into local optimization solutions due to the reduction of population diversity in the late iterations. As a consequence, Logistic algorithm is introduced to optimize SSA, making it more appropriate for solving specific problems. Upon comparative analysis, the model predicted using the Logistic-SSA-ELM algorithm was more accurate. The results indicate that the predicted values of prediction set coefficient of determination (RP), prediction set root mean square error (RMSEP) and prediction set ratio performance deviation (RPDP) were 0.7016, 3.7150 and 1.4035, respectively. This algorithm has excellent prediction performance and can be used as a model for detection of foxtail millet moisture content. In view of the important role of foxtail millet moisture content detection in acquisition and storage, it is particularly important to study a nondestructive and fast online real-time detection method. The designed capacitive sensor with differential structure has well stabilization and high accuracy, which can be further studied in depth and gradually move towards the general trend of agricultural development of smart agriculture and precision agriculture.
RESUMEN
Plant diseases can inflict varying degrees of damage on agricultural production. Therefore, identifying a rapid, non-destructive early diagnostic method is crucial for safeguarding plants. Cladosporium fulvum (C. fulvum) is one of the major diseases in tomato growth. This work presents a method of data fusion using two hyperspectral imaging systems of visible/near-infrared (VIS/NIR) and near-infrared (NIR) spectroscopy for the early diagnosis of C. fulvum in greenhouse tomatoes. First, hyperspectral images of samples at health and different times of infection were collected. The average spectral data of the image regions of interest were extracted and preprocessed for subsequent spectral datasets. Then different classification models were established for VIS/NIR and NIR data, optimized through various variable selection and data fusion methods. The principal component analysis-radial basis function neural network (PCA-RBF) model established using low-level data fusion achieved optimal results, achieving accuracies of 100% and 99.3% for calibration and prediction, respectively. Moreover, both the macro-averaged F1 (Macro-F1) values reached 1, and the geometric mean (G-mean) values reached 1 and 1, respectively. The results indicated that it was feasible to establish a PCA-RBF model by using the hyperspectral technique with low-level data fusion for the early detection of C. fulvum in greenhouse tomatoes.