Banana Pseudostem Width Detection Based on Kinect V2 Depth Sensor.

This study used Kinect V2 sensor to collect the three-dimensional point cloud data of banana pseudostem and developed an automatic measurement method of banana pseudostem width. The banana plant was selected as the research object in a banana plantation in Fusui, Guangxi. The mobile measurement of banana pseudostem was carried out at a distance of 1 m from the banana plant using the field operation platform with Kinect V2 as the collection equipment. To eliminate the background data and improve the processing speed, a cascade classifier was used to recognize banana pseudostems from the depth image, extract the region of interest (ROI), and transform the ROI into a color point cloud combined with the color image; secondly, the point cloud was sparse by down-sampling; then, the point cloud noise was removed according to the classification of large-scale and small-scale noise; finally, the stem point cloud was segmented along the y-axis, and the difference between the maximum and minimum values in the x-axis direction of each segment was calculated as its horizontal width. The center point of each segment point cloud was used to fit the slope of the stem centerline, and the average horizontal width was corrected to the stem diameter. The test results show that the average measurement error is only 2.7 mm, the average relative error was 1.34%, and the measurement time is only about 300 ms. It could provide an effective solution for the automatic and rapid measurement of stem width of banana plants and other similar plants.

Detection of chlorophyll fluorescence parameters of potato leaves based on continuous wavelet transform and spectral analysis.

The tuber development and nutrient transportation of potato crops are closely related to canopy photosynthesis dynamics. Chlorophyll fluorescence parameters of photosystem II, especially the maximum quantum yield of primary photochemistry (Fv/Fm), are intrinsic indicators for plant photosynthesis. Rapid detection of Fv/Fm of leaves by spectroscopy method instead of time-consuming pulse amplitude modulation technique could help to indicate potato development dynamics and guide field management. Accordingly, this study aims to extract fluorescence signals from hyperspectral reflectance to detect Fv/Fm. Hyperspectral imaging system and closed chlorophyll fluorescence imaging system were applied to collect the spectral data and values of Fv/Fm of 176 samples. The spectral data were decomposed by continuous wavelet transform (CWT) to obtain wavelet coefficients (WFs). Three mother wavelet functions including second derivative of Gaussian (gaus2), biorthogonal 3.3 (bior3.3) and reverse biorthogonal 3.3 (rbio3.3) were compared and the bior3.3 showed the best correlation with Fv/Fm. Two variable selection algorithms were used to select sensitive WFs of Fv/Fm including Monte Carlo uninformative variables elimination (MC-UVE) algorithm and random frog (RF) algorithm. Then the partial least squares (PLS) regression was used to establish detection models, which were labeled as bior3.3-MC-UVE-PLS and bior3.3-RF-PLS, respectively. The determination coefficients of prediction set of bior3.3-MC-UVE-PLS and bior3.3-RF-PLS were 0.8071 and 0.8218, respectively, and the root mean square errors of prediction set were 0.0181 and 0.0174, respectively. The bior3.3-RF-PLS had the best detection performance and the corresponding WFs were mainly distributed in the bands affected by fluorescence emission (650-800 nm), chlorophyll absorption and reflection. Overall, this study demonstrated the potential of CWT in fluorescence signals extraction and can serve as a guide in the quick detection of chlorophyll fluorescence parameters.

Optimal fertigation for high yield and fruit quality of greenhouse strawberry.

Nitrogen (N), phosphorus (P), potassium (K), and water are four crucial factors that have significant effects on strawberry yield and fruit quality. We used a 11 that involved 36 treatments with five levels of each of the four variables (N, P, and K fertilizers and water) to optimize fertilization and water combination for high yield and quality. Moreover, we used the SSC/TA ratio (the ratio of soluble solid content to titratable acid) as index of quality. Results showed that N fertilizer was the most important factor, followed by water and P fertilizer, and the N fertilizer had significant effect on yield and SSC/TA ratio. By contrast, the K fertilizer had significant effect only on yield. N×K fertilizer interacted significantly on yield, whereas the other interactions among the four factors had no significant effects on yield or SSC/TA ratio. The effects of the four factors on yield and SSC/TA ratio were ranked as N fertilizer > water > K fertilizer > P fertilizer and N fertilizer > P fertilizer > water > K fertilizer, respectively. The yield and SSC/TA ratio increased when NPK fertilizer and water increased, but then decreased when excessive NPK fertilizer and water were applied. The optimal fertilizer and water combination were 22.28-24.61 g plant-1 Ca (NO3)2·4H2O, 1.75-2.03 g plant-1 NaH2PO4, 12.41-13.91 g plant-1 K2SO4, and 12.00-13.05 L water plant-1 for yields of more than 110 g plant-1 and optimal SSC/TA ratio of 8.5-14.

[Detection of onion soluble solids content based on the near-infrared reflectance spectra].

Onion soluble solids content (SSC) was detected using near-infrared (924-1720 nm) reflectance spectra. Three cultivars of onions, harvested at different period, were selected for experiment and the total number of samples is 268. SSC reference value of onion juice was determined using the temperature compensated refractometer. Some pre-processing methods, such as S-G smoothing, scatter correction, and derivation, were compared to establish a statistical model based on partial least squares regression (PLSR) method. The results show that the avitzky-Golay smoothing with window 32 and span 10 is more efficient. The determination correlation coefficient of prediction R2 is 0.87 and root mean square error (RMSEP) is 2.42 degrees Brix. Compared to the 2nd derivation, the 1st derivation got better prediction result, but the spectra scatter correction is the best (R2 = 0.88, RMSEP of = 2.31 degrees Brix). The optimal prediction (R2 = 0.90, RMSEP = 1.84 degrees Brix and RPD = 3) was built based on crossing validation modeling, which shows that infrared reflectance spectroscopy with scatter correction pre-processing is feasible for onions soluble solids detection.

[Analysis and estimation of the phosphorus content in cucumber leaf in greenhouse by spectroscopy].

A handheld spectroradiometer was used to measure the spectral reflectance of the crop with the measurable range from 325 nm to 1075 nm. Since the first derivative of the spectra can well eliminate spectral error, it was calculated for each spectrum. The cucumber leaves were also sampled and the phosphorus content was measured for each sample with chemical method. First, the correlation between the phosphorus content of the cucumber leaf and the spectral reflectance was analyzed but high coefficient was not obtained. It was shown that there is not high linear relation between those. Then, the analysis was conducted between the phosphorus content of the cucumber leaf and the first derivative of spectrum for each sample. The coefficients were improved. However, it was not high enough to establish an estimation model. It shows that non-linear model is needed to estimate the phosphorus content of the crop leaf based on spectral reflectance. Artificial neural network (ANN) and support vector machine (SVM), the modern ealgorithm for modeling and estimating, were used to establish the nonlinear models. From stepwise multi-regression, four wavelengths, 978, 920, 737 and 458 nm, were selected as modeling wavebands. For the Artificial Neural Network (ANN) model, the data of spectral reflectance in the four wavebands were taken as the input and the phosphorus content was taken as the output. And the number of the neurons in the middle layer, the learning rate, and the learning error were set as 25, 0.05, and 0.001, respectively. The calibration accuracy of the model was 0.995, and the validation accuracy reached to 0.712. For the Support Vector Machine (SVM) model, the selected kernel function was anova, and the penalty parameter C and the linear epsilon-insensitive loss function were set as 100 and 0.00001, respectively. The calibration accuracy of the model was closed to 1, and the validation accuracy reached to 0.754. It can be concluded that both nonlinear models are practical.