Regular article
Measurement orientation compensation and comparison of transmission spectroscopy for online detection of moldy apple core

https://doi.org/10.1016/j.infrared.2020.103510Get rights and content

Highlights

  • Develop an online system for moldy core detection.

  • Stability study of spectra acquired in three fruit orientations using SNR and ACR.

  • Analysis of spectral propagation in different orientations.

  • Comparison to moldy core detection models of local and global orientations.

Abstract

Moldy core is an internal disease afflicting otherwise healthy apples. This paper reports the results of a study on the effect of fruit orientation on the online detection of moldy core in Fuji apples using Vis/NIR transmission spectroscopy. Using a self-designed online spectral measurement system, transmission spectra in the range 550–900 nm were collected from apples positioned in three orientations: vertical calyx-stem axis with stem upward (T1), horizontal calyx-stem axis (T2), and vertical calyx-stem axis with stem downward (T3). In order to evaluate the spectral stability of the system in different orientations, the area change rate (ACR) and signal-to-noise ratio (SNR) were introduced. The propagation of light inside the fruit was also analyzed. After excluding outliers, 244 apple samples were divided into a training set (113 healthy apples and 70 diseased apples) and a test set (38 healthy apples and 23 diseased apples) at a ratio of 3:1. After spectral pretreatment, a support vector machine (SVM) was used to develop compensation models for the detection of moldy core in apples at all orientations (global model) and each separate orientation (local models). The results show that T2 provides the best performance in our system, and that the measurement orientation significantly affects the recognition accuracy of moldy core in apples. In contrast, the global model is insensitive to changes in fruit orientation, and the recognition accuracy of the global SVM model is 100% in all three orientations. Thus, it is considered that the global SVM model is a promising candidate for the online detection of moldy core in apples using transmission spectroscopy, regardless of the spectral measurement orientation.

Introduction

Apples are widely available in the global market and are popular around the world for their excellent flavor and nutritional value [1], [2]. The ever-increasing consumer awareness of food quality makes it essential for the apple industry to maintain high standards of fruit quality [3], especially because the revenue from such high-quality products is much higher than the average income. However, apples are susceptible to a variety of defects in its pre-harvest, at harvest and during post-harvest operations, such as storage, transport and retail distribution. Therefore, it is important to detect these defects after harvest and before sale.

The defects suffered by apples generally fall into the following categories: physiological disorders, mechanical damage, pathological disorders, and contamination [1]. Pathological disorders and contamination usually cause more serious defects than the other categories, and these defects have seriously endangered food safety, so fruit inspection for these issues is performed using higher and stricter standards. Because of the acidity of the apple tissue, apples are susceptible to fungal infection, which can cause pathological disorders [4], [5], [6], [7], [8]. One of the most common pathological disorders is moldy core, which is characterized by dry, spongy brown lesions in the core area [9], [10], [11]. Preventing such defects is very difficult, and apples with moldy cores cannot be detected until the apples are cut open [12]. Ideally, diseased fruit would be removed prior to storage; otherwise, healthy apples are likely to be infected. Therefore, there is a strong need within the apple industry for a reliable, nondestructive method for detecting and grading such fruit.

Several nondestructive methods for detecting internal defects in apples have been developed. These methods involve X-ray imaging [13], [14], magnetic resonance imaging [15], [16], [17], [18], and thermal imaging [19], but the commercial application of these methods has a big drawback, that is, they are very expensive and time-consuming. Vis/NIR transmission spectroscopy is a fast, cheap, and nondestructive method for the detection of internal apple diseases. This technology was first used to detect internal diseases in apples in the 1980s and is now largely mature [20], [21], [22], [23], [24], [25], [26], [27]. In the research of Upchurch et al. [28], the ability of transmission spectroscopy to detect internal diseases in apples had been demonstrated. After this study, four arrangements of light source, fruit-orientation and detector used to detect the brownheart in apples were investigated by Clark et al. [29]. Guo et al. [30] quantitatively analyzed the internal quality and diseases in apples by using near infrared transmittance spectroscopy and obtained good detection results. These studies focused on the static, rather than the dynamic or online, state of fruit. For commercial applications, this type of research is insufficient, and a technology that can rapidly detect moldy core in an online manner, rather than statically, is urgently required. Shenderey et al. [31] evaluated the ability of Vis-NIR minispectrometers to detect moldy core in apples, on line. This method provides a research basis for online detection in commercial applications. McGlone et al. [32] developed two prototype online transmission spectroscopy systems. The influence of two different spectrometers on the detection results has been investigated in the literature. However, measurement orientation is also a significant factor affecting the quality of the acquired spectra during online detection [33], [34]. Fu et al. [35] used transmission spectra to detect brown heart in pears, and better results were obtained in fruit orientation T2. Han et al. [36] compared three fruit orientations (T1, T2, T3) for detecting the soluble solids content in apples, and obtained better results in fruit orientation T1. Clearly, during the online detection of moldy core in apples, the methods to evaluate the spectral information in different orientations and the methods to eliminate unwanted effects are lacking, and it is necessary for us to study this part of the content.

The general goal of this study was to develop an optimal model to realize the online detection of moldy core in Fuji apples, in terms of fruit orientation. The specific objectives were (1) to investigate the performance of an online system that collects the transmission spectra of apples; (2) to evaluate the stability of spectral information obtained from different orientations using the signal-to-noise ratio (SNR) and area change rate (ACR); and (3) to establish and compare the moldy core detection models generated by the local orientation and the global orientation.

Section snippets

Samples

To better describe our samples, we refer to the apples with a moldy core as diseased apples and the apples with a normal core as healthy apples (Fig. 1). The research object were red Fuji apples. Using portable equipment developed in our laboratory [37], 132 suspected diseased apples and 123 suspected healthy apples were selected for subsequent experimental analysis. All 255 Fuji apples were collected in late September 2019 from the same refrigerated storage (37°25′5.63″N, 120°53′5.84″E, 82 m

Overview of spectra and sample data distribution

The raw transmission spectrum obtained by the system in the T1 orientation was shown in Fig. 5. Observations of the spectrum revealed two distinct absorption peaks at 640 nm and 714 nm. The spectra obtained in orientations T2 and T3 had similar characteristics, and their spectral variation trends were also the same. However, such spectra are useless for detecting moldy core because they contain substantial dark noise information and baseline shifts. The method based on principal component

Conclusions

The influence of variations in spectral measurement orientation on the online detection of moldy core in apples using transmission spectroscopy was studied using a self-designed online spectral measurement system. Through SNR and ANR analysis, we concluded that the spectral information obtained by the system in the T2 orientation was the most stable. After obtaining spectral information in orientations T1, T2, and T3, local and global SVM models were established after SGS and normalization

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This study was financially supported by the Key Research and Development Program of Shaanxi (Item No. 2017ZDXM-NY-017), the National Natural Science Foundation of China (Grant No. 31701664), the Horizontal Scientific Research Project (Item No. ZY-XG-CG-2019-07-06), and the Science and Technology Coordination Innovation Project of Shaanxi (Item No. 2016KTCQ02-14) www.liwenbianji.cn/ac.

References (50)

  • J.A. Abbott

    Quality measurement of fruits and vegetables

    Postharvest Biol. Technol.

    (1999)
  • P. Chen et al.

    A review of non-destructive methods for quality evaluation and sorting of agricultural products

    J. Agr. Eng. Res.

    (1991)
  • B.M. Nicolai et al.

    Nondestructive measurement of fruit and vegetable quality by means of NIR spectroscopy: a review

    Postharvest Biol. Technol.

    (2007)
  • T. Ringsted et al.

    Long wavelength near-infrared transmission spectroscopy of barley seeds using a supercontinuum laser: prediction of mixed-linkage beta-glucan content

    Anal. Chim. Acta

    (2017)
  • Y. Xia et al.

    Effect of spectral measurement orientation on online prediction of soluble solids content of apple using Vis/NIR diffuse reflectance

    Infrared Phys. Technol.

    (2019)
  • B.L. Upchurch et al.

    Detecting internal breakdown in apples using interactance measurements

    Postharvest Biol. Technol.

    (1997)
  • C.J. Clark et al.

    Detection of Brownheart in ‘Braeburn’ apple by transmission NIR spectroscopy

    Postharvest Biol. Technol.

    (2003)
  • Z. Guo et al.

    Quantitative detection of apple watercore and soluble solids content by near infrared transmittance spectroscopy

    J. Food Eng.

    (2020)
  • V.A. McGlone et al.

    On-line detection of Brownheart in Braeburn apples using near infrared transmission measurements

    Postharvest Biol. Technol.

    (2005)
  • G.Q. Fan et al.

    Determination of soluble solids and firmness of apples by Vis/NIR transmittance

    J. Food Eng.

    (2009)
  • S.X. Fan et al.

    Effect of spectrum measurement position variation on the robustness of NIR spectroscopy models for soluble solids content of apple

    Biosyst. Eng.

    (2016)
  • X.P. Fu et al.

    Comparison of diffuse reflectance and transmission mode of visible-near infrared spectroscopy for detecting brown heart of pear

    J. Food Eng.

    (2007)
  • M. Ruiz-Altisent et al.

    Instrumental quality assessment of peaches: fusion of optical and mechanical parameters

    J. Food Eng.

    (2006)
  • L.R. Zhang et al.

    Use of signal to noise ratio and area change rate of spectra to evaluate the Visible/NIR spectral system for fruit internal quality detection

    J. Food Eng.

    (2014)
  • R.K.H. Galvao et al.

    A method for calibration and validation subset partitioning

    Talanta

    (2005)
  • Cited by (0)

    View full text