Baohua Zhang

Hyperspectral imaging combined with multivariate analysis and band math for detection of common defects on peaches (Prunus persica)

Common defects are divided into two different types.For artificial defects, multivariate analysis were used for wavelengths selection.For non-artificial defect and stem detection, band math were constructed.The uneven lightness distribution was also investigated in this paper. Automatic detection of common defects on peaches by using imaging system is still a challenge due to the high variability of peach surface color, the similarity between the defects and stem, as well as the uneven distribution of lightness on peaches. In order to detect the common defects on peaches using hyperspectral imaging, defects were divided into two different types: artificial defects and non-artificial defects. For artificial defect detection, a two-step multivariate analysis method (Monte Carlo-Uninformative Variable Elimination and successful projections algorithm) was conducted in the spectral domain for the discriminant wavelength (DW) selection, and then minimum noise fraction (MNF) transform was conducted on the images at DWs for image processing and artificial defect detection. For the candidate non-artificial defect detection, a pair of two characteristic wavelengths at 925nm and 726nm was selected by analyzing the full spectra of sound and non-artificial defective regions, and then a band math equation was constructed for differentiating the non-artificial defect regions and stems from the sound and physical damage regions, and the candidate non-artificial defects (including non-artificial defects and stems) could be segmented by using a simple threshold method. In order to distinguish the stem from the segmented candidate non-artificial defect regions, another band math equation was constructed based on another pair of two characteristic wavelengths at 650nm and 675nm for stem identification. Additionally, the uneven lightness distribution in the spectral images was also investigated and eliminated by the band math methods. The overall classification accuracy of 93.3% for the 120 samples indicated that the selected DWs and proposed method were suitable and efficient for the common defect detection. The limitation of our research is the static inspection in one single view.

Variable Selection in Visible and Near-Infrared Spectral Analysis for Noninvasive Determination of Soluble Solids Content of ‘Ya’ Pear

Informative variable selection or wavelength selection plays an important role in the quantitative analysis of near-infrared (NIR) spectra because the modern spectroscopy instrumentations usually have a high resolution and the obtained spectral data sets may have thousands of variables and hundreds or thousands of samples. In this study, a new combination of Monte Carlo–uninformative variable elimination (MC-UVE) and successive projections algorithm (SPA; MC-UVE-SPA) was proposed to select the most effective variables. MC-UVE was firstly used to eliminate the uninformative variables in the raw spectra data. Then, SPA was applied to determine the variables with the least collinearity. A case study was done based on the NIR spectroscopy for the non-destructive determination of soluble solids content (SSC) in ‘Ya’ pear. A total of 160 samples were prepared for the calibration (n = 120) and prediction (n = 40) sets. Three calibration algorithms including linear regressions of partial least square regression (PLS) and multiple linear regression (MLR), and nonlinear regression of least-square support vector machine (LS-SVM) were used for model establishment by using the selected variables by SPA, UVE, MC-UVE, UVE-SPA, and MC-UVE-SPA, respectively. The results indicated that linear models such as PLS and MLR were more effective than nonlinear model such as LS-SVM in the prediction of SSC of ‘Ya’ pear. In terms of linear models, different variable selection methods can obtain a similar result with the RMSEP values range from 0.2437 to 0.2830. However, combination of MC-UVE and SPA was helpful for obtaining a more parsimonious and efficient model for predicting the SSC values in ‘Ya’ pear. Twenty-two effective variables selected by MC-UVE-SPA achieved the optimal linear MC-UVE-SPA-MLR model compared with other all developed models by balancing between model accuracy and model complexity. The coefficients of determination (r2), root mean square error of prediction, and residual predictive deviation by MC-UVE-SPA-MLR were 0.9271, 0.2522, and 3.7037, respectively.

Design and Implementation of an Automatic Grading System of Diced Potatoes Based on Machine Vision

Potato is one of the most important crops in the world. In recent years, potato and its processed products have gradually become important trade goods. As an important semi-manufactured product, diced potatoes need to be graded according to their three-dimensional (3D) size and shape before trading. 3D information inspection manually is a time-consuming and labor intensive work. A novel automatic grading system based on computer vision and near-infrared linear-array structured lighting was proposed in this paper. Two-dimensional size and shape information were extracted from RGB images, and height information was measured in NIR images combined with structured lighting. Then, a pair of pseudo-color and gray level height map images fusing with 3D size and shape information was constructed. Finally, diced potatoes were classified into either regular or irregular class according to their 3D information and criteria required by the industry. The grading system and proposed algorithm were testified by a total of 400 diced potatoes with different size and shapes. The test results showed that the detection error was in the range of about 1 mm, and the classification accuracy was 98 %. The results indicated that the system and algorithm was efficient and suitable for the 3D characteristic inspection of diced potatoes.

Quality of Vegetable Products: Assessment of Physical, Chemical, and Microbiological Changes in Vegetable Products by Nondestructive Methods

This chapter focuses on quality and safety evaluation of the vegetable products using different sensing technologies, imaging processing, and chemometric methods. It provides an overview of the instruments used for evaluating the quality of vegetable products such as computer vision, multispectral imaging, near-infrared spectroscopy, and hyperspectral imaging (refer to Sect. 2). Then, the basic analysis methods and chemometrics are introduced in detail (Sect. 3), including image/spectral preprocessing and correction/calibration, feature and band extraction and sample selection, and analysis models and evaluation. Finally, the potential applications of the instruments and the basic analysis methods in vegetable product quality and safety analysis and control are explained (Sect. 4). The external qualities such as shape, size, color, texture, and defects; internal qualities such as soluble solid content (SSC), acid content, and internal defects; and microbiological changes such as microbial and fecal contamination are discussed in detail. Conclusions and future works are proposed (Sect. 5).

Penetration Depth of Near-Infrared Light in Small, Thin-Skin Watermelon

Non-destructive detection of internal quality in watermelon has very important significance for improving watermelon’s production efficiency. Near-infrared (NIR) spectroscopy is one of the most popular non-destructive detection methods. However, it is challenging to collect spectra exactly due to the multiple scattering and absorbing by the skin and internal tissues. In order to obtain the interactions between light and watermelon tissues, the transportation feature of NIR light in small, thin-skin watermelon was studied in the range of 750–900 nm. For this purpose, the diffused transmission spectra were collected with removing the sample slices along the perpendicular bisector of the source-detector line. Based on the spectra in effective wavelength band, the penetration depth curves were fitted by least square method, and the results of different detecting positions (equator and top) were compared. It was shown that, light penetration depth on the equator was 8.3–9.5 mm, 8.7–17.8 mm and 18.9–38.5 mm with source-detector distance of 10 mm, 20 mm and 30 mm, respectively. The penetration depth on the top was less than the equator. And the penetration depth increased with source-detector distance increasing. With deeper penetration depth, more information about internal quality was carried by the diffused transmission spectra. However, the intensity of spectra was weaker. According to these results, a reasonable source-detector distance could be designed for collecting effective information about internal quality. This study is of potential significance for optimizing the handheld probe geometry for large fruit, and offers theoretical bases for non-destructive detection.

Comparison of Four Types of Raman Spectroscopy for Noninvasive Determination of Carotenoids in Agricultural Products

Carotenoids are one class of naturally-occurring pigments with antioxidant properties. They can absorb light energy for use in photosynthesis for plants, and act as antioxidants to reduce risk of cancer for human. Carotenoids are confirmed to exist in agricultural products as the main source for human. Raman spectroscopy is a new technique for determination of carotenoids in agricultural products as it is both noninvasive and rapid. Four types of Raman spectroscopy could be used for contact measurement of carotenoids in fruits and vegetables: (1) Fourier transform Raman spectroscopy; (2) Resonance Raman spectroscopy; (3) Raman microspectroscopy; (4) Spatially Offset Raman spectroscopy. The experimental setups, advantages and applications of the above-mentioned Raman spectroscopies are discussed.