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

Abstract

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%.

Applying hyperspectral imaging to explore natural plant diversity towards improving salt stress tolerance

“…Application of HSI was done for food quality evaluation soybean seeds, wheat, barley grains and Portobello mushroom (Agaricus bisporu) (Taghizadeh et al, 2011, Tumuluru et al, 2010, Arngren et al, 2011, Huang et al, 2014a, 2014b and detection of defects in apple fruits and lettuce leaves (Baranowski et al, 2012, Simko et al, 2015. In parallel, experiments using HSI techniques for detection of fungal infections in fruits of citrus, leaves of sugar beet, wheat and maize have proven the applicability of the technique (Lorente et al, 2013, Mahlein et al, 2010Hillnhütter et al, 2011, Firrao et al, 2010Yao et al, 2010, Bauriegel and Herppich, 2014, Williams et al, 2012. Nowadays, HSI methodologies based on various indices have been mostly used for detection of photosynthetic pigments like chlorophylls, carotenoids as well as of the other major compounds in leaves and fruits, such as anthocyanins and flavonoids (Deepak et al, 2015;Matros and Mock, 2013, Hölscher et al, 2009, Zhao et al, 2005, Coops et al, 2003, Ferri et al, 2004.…”

Automatic Robust Segmentation Scheme for Pathological Problems in Mango Crop

“…In [22] the authosr have calculated Optimal Wavelength Features using neural network and ROC (Receiver Operating Characteristic) curves in order to determine the Decay for Citrus Fruit. Here, receiver operating curves are used in order to choose features in multiclass classification issues where with the use of hyper spectral images decay is detected in citrus fruits.…”

Automated Systems Based on Machine Vision for Inspecting Citrus Fruits from the Field to Postharvest—a Review

“…These techniques transform the data in the high-dimensional space into a lower-dimensional space that preserves the observed properties of the data. The spectral features are used as inputs of classification algorithms in order to increase the performance of the classifiers developed to discriminate between sound and decaying skin (Gomez-Sanchis et al 2013;Lorente et al 2013aLorente et al & 2013b.…”