Integrating multispectral reflectance and fluorescence imaging for defect detection on apples

Ariana, Diwan P.; Guyer, Daniel E.; Shrestha, B

doi:10.1016/j.compag.2005.10.002

Cited by 98 publications

(44 citation statements)

References 29 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…By considering a significant result in all three experiments, the best detection was provided using F 440 /F 740 which increased in botrytised berries when compared to control ones. The earliest detection made using this fluorescence ratio was achieved at 4 DAI in experiment C. Ariana et al (2006) also observed an increase in F 440 /F 740 for black rot tissue on apples compared to healthy ones. Fluorescence ratios not involving blue-green fluorescence (from 440 nm to 520 nm) did not achieve successful detection of Botrytis using a spatial average over the berry area.…”

Section: Fluorescence Data Based On Spatial Average Over the Berry Areamentioning

confidence: 81%

Autofluorescence of grape berries following Botrytis cinerea infection

Bélanger¹,

Roger²,

Cartolaro

et al. 2011

International Journal of Remote Sensing

View full text Add to dashboard Cite

ξ current address : Financière agricole du Québec -Développement international, 1400, boulevard de la Rive-Sud, Saint-Romuald, Québec, G6W 8K7, Canada Gray mold is caused by Botrytis cinerea (anamorph of an ascomycete fungus) infecting over 200 plant species worldwide and causing tremendous harvest losses in vineyards. Even though all grapevine cultivars (Vitis vinfera L.) are susceptible to the disease, defense mechanisms are induced to counteract or slow down infection and colonization by the pathogen. One of the key inducible defense molecule is resveratrol, a blue fluorescent stilbenic compound. Considering early fungal detection as a crucial step in disease control strategies, we evaluated the potential of plant fluorescence to achieve early detection of B.cinerea in grapevine berries. By comparing control and artificially inoculated berries, images of fluorescence were taken in the laboratory at six combinations of emission/excitation wavelengths (F440/ UV , F520/ UV , F690/ UV, F740/ UV , and F690/436-blue F740/436-blue). Data treatments include proper image correction, interest area selection, fluorescence ratios computation and edge detection. In the early stages, before 3 DAI (days post inoculation), the discrimination (of -between ?) wounded and symptomatic berries could not easily be achieved using blue fluorescence solely (440 nm). However, using blue out of red Author-produced version of the article published in International Journal of Remote Sensing, 2011, 32 (14), 3835-3849. The original publicatioin is available at http://www.tandfonline.com doi:10.1080/01431161003782064Page 2 fluorescence ratio (F 440 /F 740 ), it was possible to detect infected berries starting as early as 4DAI. Using image analysis and edge detection over UV-epidermal transmittance measured at 690 nm, it was possible to detect botrytised berries as early as 6 hours after inoculation. The potential of these detection methods for use in the epidemiological field are discussed.Key words: Bunch rot, detection, epidemiology, fluorescence, remote sensing, phytoalexins, precision viticulture, Vitis vinifera. IntroductionGray mold of grapevines is caused by Botrytis cinerea Pers.:Fr., the anamorph of an ascomycete fungus (Botryotinia fuckeliana Whetzel). This cosmopolitan fungus is a necrotrophic pathogen attacking over 200 plant species (Jarvis, 1980). In grapevine, it can infect leaves, in spring, where it grows in a concentric way at the margin of the blade and inflorescences which are highly susceptible, particularly dehiscent floral parts (Pezet et al. 2004). Various infection pathways have been described to establish epidemiological relationships between these first attacks, occurring early in the season, and the typical infections of maturing berries (Elmer and Michailides, 2004). Gray mold development on grape berries depends on the genetic structure of the pathogen population (Martinez et al., 2005), but is also driven by some key factors, including climatic conditions, cluster architecture and berry susceptibility (Pezet...

show abstract

Section: Fluorescence Data Based On Spatial Average Over the Berry Areamentioning

confidence: 81%

Autofluorescence of grape berries following Botrytis cinerea infection

Bélanger¹,

Roger²,

Cartolaro

et al. 2011

International Journal of Remote Sensing

View full text Add to dashboard Cite

show abstract

“…Researchers used different sensing techniques like X-ray imaging (Diener et al, 1970;Shahin et al, 2002), hyperspectral imaging (Lu, 2003;Mehl et al, 2004;ElMasry et al, 2009;Sun, 2010) and spectral reflectance based methods (Upchurch et al, 1991 ;Geoola et al, 1994;Ariana et al, 2006;Peng and Lu, 2008) to grade apple fruit. However, majority of the works for this problem include systems based on visible/near infrared (NIR) imaging, which can be divided into two sub-groups: (1) those employing special equipment and (2) those using ordinary machine vision.…”

Section: Introductionmentioning

confidence: 99%

Automatic grading of Bi-colored apples by multispectral machine vision

Ünay

Gosselin

Kleynen

et al. 2011

Computers and Electronics in Agriculture

134

View full text Add to dashboard Cite

In this paper we present a novel application work for grading of apple fruits by machine vision. Following precise segmentation of defects by minimal confusion with stem/calyx areas on multispectral images, statistical, textural and geometric features are extracted from the segmented area. Using these features, statistical and syntactical classifiers are trained for two-and multi-category grading of the fruits. Results showed that feature selection provided improved performance by retaining only the important features, and statistical classifiers outperformed their syntactical counterparts. Compared to the state-of-the-art, our two-category grading solution achieved better recognition rates (93.5% overall accuracy). In this work we further provided a more realistic multi-category grading solution, where different classification architectures are evaluated. Our observations showed that the single-classifier architecture is computationally less demanding, while the cascaded one is more accurate.

show abstract

“…Unay and Gosselin (2006) used a multilayer perceptron (MLP) as a promising technique for segmenting surface defects on apples. Ariana et al (2006) presented an integrated approach using multispectral imaging in reflectance and fluorescence modes to acquire images of three varieties using two ANN-based classification schemes (binary and multi-class). In the case of citrus fruits, Kondo et al (2000) used, among other methods, ANN to detect some external and internal features in oranges while Gómez-Sanchis et al (2012) used minimum redundancy maximum relevance as feature selection method and MLP for pixel classification to detect rottenness in mandarins.…”

Section: Introductionmentioning

confidence: 99%

Selection of Optimal Wavelength Features for Decay Detection in Citrus Fruit Using the ROC Curve and Neural Networks

Lorente

Aleixos

Gómez-Sanchís

et al. 2011

Food Bioprocess Technol

118

View full text Add to dashboard Cite

Early automatic detection of fungal infections in post-harvest citrus fruits is especially important for the citrus industry because only a few infected fruits can spread the infection to a whole batch during operations such as storage or exportation, thus causing great economic losses. Nowadays, this detection is carried out manually by trained workers illuminating the fruit with dangerous ultraviolet lighting. The use of hyperspectral imaging systems makes it possible to advance in the development of systems capable of carrying out this detection process automatically. However, these systems present the disadvantage of generating a huge amount of data, which must be selected in order to achieve a result that is useful to the sector. This work proposes a methodology to select features in multiclass classification problems using the receiver operating characteristic curve, in order to detect rottenness in citrus fruits by means of hyperspectral images. The classifier used is a multilayer perceptron, trained with a new learning algorithm called extreme learning machine. The results are obtained using images of mandarins with the pixels labelled in five different classes: two kinds of sound skin, two kinds of decay and scars. This method yields a reduced set of features and a classification success rate of around 89%.

show abstract

Integrating multispectral reflectance and fluorescence imaging for defect detection on apples

Cited by 98 publications

References 29 publications

Autofluorescence of grape berries following Botrytis cinerea infection

Autofluorescence of grape berries following Botrytis cinerea infection

Automatic grading of Bi-colored apples by multispectral machine vision

Selection of Optimal Wavelength Features for Decay Detection in Citrus Fruit Using the ROC Curve and Neural Networks

Contact Info

Product

Resources

About