Hyperspectral imaging (HSI) is one of the most studied optical techniques to estimate the internal quality of fruits and vegetables. Absorbance and reflectance of the light radiation are specific to each biological tissue and are directly related to its chemical composition and physical characteristics. These properties are influenced by other extrinsic factors, such as the instrumentation or the light source, which can reduce the reproducibility of the experiments. Determining the actual depth of light penetration into tissue could help validate non-contact methods as accurate tools to assess quality properties based on optical properties. In the case of HSI systems, it is crucial to know how far the light penetrates at each wavelength. A non-destructive approach, based on the spatially resolved spectroscopic principle, was proposed to estimate the light penetration depth of a HSI system in a Vis-NIR configuration (in the range 450-1050 nm). This method was applied to measure the light penetration depth in persimmon fruit. The absorption (mu(a)) and scattering (mu'(s)) coefficients from Farrell's diffusion theory were estimated using the backscattered light measured at different distances from the incident point light at each wavelength in hyperspectral images of persimmon fruit. The actual light penetration depth was obtained by measuring the reflectance of cut pieces of persimmon fruit with different thicknesses. Linear regression was used to relate the depth of penetrability obtained by both protocols, the estimated or non-destructive protocol and the actual or destructive protocol, showing a high relationship (R-2 > 0.8 and RPD>2.5) in the range 610-1050 nm. This confirms that this non-destructive approach proposed for estimating the light penetration depth of a Vis-NIR HSI system in persimmon fruit is accurate, so it could be used as a valuable method to evaluate other HSI systems for different fruits.

The collection of oranges normally begins before they have reached the typical orange colour. Moreover, citrus fruits are subjected to certain degreening treatments that depend on the standard citrus colour index (CCI) at harvest. In order to facilitate the measure of this index, a free application that uses image processing techniques has been developed for Android-based mobile devices using the built-in camera of the device. The image analysis process is performed on all the images from the live input of the camera to obtain the CCI of such fruit using the open source OpenCV library. For this purpose, the RGB (red, green and blue colour coordinates) average value of a pre-selected area of the input image is calculated and then converted to HunterLab colour space to finally calculate the CCI. Several tests were carried out in the field with the fruit on the trees and under laboratory conditions with different varieties of oranges (Navel, Bonanza, Cram and Navelina) at different stages of maturity, and using different Android devices. The results were obtained for each device and condition in relation to the colour measured by a camera and compared with the performance of a panel of workers who evaluated the colour using the traditional methods. Best R-2 values obtained were 0.854 for outdoors conditions and 0.881 when measurements were done indoors. (C) 2017 lAgrE. Published by Elsevier Ltd. All rights reserved.

Nowadays the harvesting of coffee in Colombia is done manually, has a high demand for labour and is responsible for 40% of production costs. To provide hiring staff, producers make destructive or subjective estimations of the yield production. This work aims to generate an automatic strategy for planning the coffee harvest through nondestructive estimation of the percentage of mature fruits (PM) by means of image analysis. Images of 69 coffee branches of Castillo (R) variety were acquired using a mobile phone in different field conditions, having a real PM between 10% and 70%. Due to the high length of the branches (40 to 60 cm), multiple images along each branch were captured, obtaining a total of 280 images. The Food-Colorlnspector application (http://www.cofilab.com) was used to segment the images, and the area corresponding to three stages of maturity of fruits (immature, half-ripe and full-ripe) was obtained for each branch. Additionally, manual counting of the fruits at different ripening stages was performed. A total of 23 branches were used to create a linear model using the percentages of maturity obtained by image analysis and manual analysis. The model obtained a coefficient of determination R-2=88%, with an absolute average error of 3.9% between the PM estimated by hand and by image analysis. Subsequently, this model was validated with the remaining 46 branches, obtaining a R-2=75% and an average error of 5.5%. The results are promising to create tools to automate current manual yield prediction of coffee by using mobile devices such as smartphones, available to any farmer

Tangent and Corner Vertices Detection (TCVD) is a method to detect corner vertices and tangent points in sketches using,parametric cubic curves approximation, which is capable to detect corners with a high accuracy and a very low false positive rate, and also to detect tangent points far above other methods in literature. In this article, we present several improvements to TCVD method in order to establish mathematical conditions to detect corners and make the obtaining of curves independent from the scale, what increases the success ratio in transitions between lines and curves. The new conditions for obtaining corners use the radius as the inverse of the curvature, and the second derivative of the curvature. For the detection of curves, a new descriptor is presented, avoiding the parameters dependent of scale used in TCVD method.In order to obtain the performance of the implemented improvements, several tests have been carried out using a dataset which contains sketches more complex than those used for validation of TCVD algorithm (sketches with more curves and tangent points and sketches of different sizes). For corners detection, the accuracy obtained was pretty similar to that obtained with the previous TCVD, however, for curves and tangent points detection the accuracy increases significantly.

Food-Colorlnspector is a software application specially developed to evaluate the colour and some external characteristics of foods by image analysis. It allows the user to select different areas in the image and assign them to one of the predefined classes representing different regions of interest. The algorithms developed use the selected areas (training set) to create a colour map that allows segmenting the image by classifying each pixel in different predefined classes. The RGB (red, green, blue) pixel coordinates in the selected areas are used as independent variables and the class to which they belong as a dependent variable, for creating a classifier based on Bayes' theorem. Using this information, the application creates a model capable of segmenting the image by classifying any pixel in the image in one of the predefined classes. This model is saved in a colour map that contains every possible combination of RGB in an image and the class to which it is assigned. From the segmented image, basic statistics of the colour in different colour spaces (RGB, CIELAB and HSI) and the main geometric characteristics of each found region are provided. The colour map generated with the training images can be later used to process sets of images automatically, providing information for each individual image and the whole set in a spreadsheet. Among other applications, it can be used, for example, to estimate the browning of a product regardless of the surface texture, which is difficult to achieve using a traditional colorimeter. It can also be used to detect, measure and characterize external damages, stains or maturity stages or differentiate between different regions corresponding to different food composition. The application can be free downloaded from the website http:\\ www. cofilab. com well as a collection of testing images

Mango is a tropical fruit with high added value and has been interest by food industry in last years, which is increasing its efforts to determine the internal quality of this fruit by non-destructive techniques. In this study, 131 mangoes of 'Kent' variety were used and divided into three batches (unripe, ripe and overripe) and stored under controlled temperature and humidity. Mango images were taken by two sides with a hyperspectral systems based on two liquid crystal tunable filters (LCTF), sensitive in the spectral range 420-1080 nm. After images capturing, firmness, acidity and soluble solids content were analyzed by reference destructive techniques. A ripening index (RPI) was established with the combination of these properties. Hyperspectral images were analyzed by extracting the surface reflectance of both sides of mangoes. The data analysis was performed by two models of partial least squares regression (PLS) to establish the relationship between reflectance spectra of the surface of the mangoes and RPI obtained. 70% of the samples was used to build the model, obtaining a R-2 = 0,852 with all the bands for the calibration set. Later, the set was reduced to six wavelengths (520, 560, 730, 760, 870 and 1050 nm) with R-2 = 0,849. The model with six wavelengths were validated with 30% of the remaining samples to give a R-2 = 0,739 indicating that it is possible to predict the maturity of this variety of mango using a very limited set of wavelengths in the visible and near infrared.

A fruit inspection system based in real-time computer vision has been designed especially to be embarked on an agricultural vehicle, consisting of a camera, a lighting system based on LED technology and an industrial computer. The system is powered entirely by solar panels and can be incorporated into an agricultural vehicle capable of transporting fruit through an inspection chamber. The LEDs powered in strobe mode, the acquisition and analysis of the images, and the subsequent separation of the fruit are synchronized with the advance of the fruit by incremental encoder connected to the shaft of the conveyor belt. In this paper, the system was tested by inspecting oranges on a mobile platform to assist in the citrus harvesting, with capacity to individualise, transporting and separating the fruit into quality categories according to their colour, size and defects. An algorithm of image segmentation was developed to detect the most common skin blemishes of citrus fruits, the size and the colour from multiple views of the fruit acquired while it moves rotating through the inspection area. Finally, the vision system communicates with the control of the machine to separate the fruit by the outlets corresponding to their quality. Preliminary results are promising and demonstrate the potential of these systems to perform a presorting of the fruit while it is collected and detect those that do not achieve quality to be sent to the market.

Computer vision systems are becoming a scientific but also a commercial tool for food quality assessment. In the field, these systems can be used to predict yield, as well as for robotic harvesting or the early detection of potentially dangerous diseases. In postharvest handling, it is mostly used for the automated inspection of the external quality of the fruits and for sorting them into commercial categories at very high speed. More recently, the use of hyperspectral imaging is allowing the detection of not only defects in the skin of the fruits but also their association to certain diseases of particular importance. In the research works that use this technology, wavelengths that play a significant role in detecting some of these dangerous diseases are found, leading to the development of multispectral imaging systems that can be used in industry. This article reviews recent works that use colour and non-standard computer vision systems for the automated inspection of citrus. It explains the different technologies available to acquire the images and their use for the non-destructive inspection of internal and external features of these fruits. Particular attention is paid to inspection for the early detection of some dangerous diseases like citrus canker, black spot, decay or citrus Huanglongbing.

The first citrus fruits harvested in Spain before they have reached their typical orange. However, before being sent to market, it is necessary that the fruit reaches a certain commercial color, so the fruit is subjected to treatments of degreening whose duration depends on its color at harvest according to standard Citrus Color Index (ICC). To estimate this index, fruits are compared with charts representing the citrus skin with different colour indices or colorimeters are used to determine the colour. In order to facilitate the estimation of this index, a free application that uses image processing techniques has been developed for Android-based mobile devices using the built-in camera of the device. Images of a fruit are captured and analyzed in-vivo to obtain the ICC. For this, the average value of the colour coordinates (red, green, blue) of a circular spot in the image is calculated and transformed into the HunterLab color space to calculate finally the ICC. In order to validate this method, several tests were carried out in both, the field and under laboratory conditions with different varieties of oranges (Navel, Bonanza, Cram and Navelina) at different stages of maturity, and using different Android devices. The measurements obtained were compared to the results provided by a colorimeter. The R-2 values obtained ranged from 0.96 to 0.98 for the different tested mobile devices, showing the robustness and reliability of the mobile-based application developed to be used to assist in the decision making about the harvesting time or treatment of these fruits.

This article presents a new method for analysing mosaics based on the mathematical principles of Symmetry Groups. This method has been developed to get the understanding present in patterns by extracting the objects that form them, their lattice, and the Wallpaper Group. The main novelty of this method resides in the creation of a higher level of knowledge based on objects, which makes it possible to classify the objects, to extract their main features (Point Group, principal axes, etc.), and the relationships between them. In order to validate the method, several tests were carried out on a set of Islamic Geometric Patterns from different sources, for which the Wallpaper Group has been successfully obtained in 85% of the cases. This method can be applied to any kind of pattern that presents a Wallpaper Group. Possible applications of this computational method include pattern classification, cataloguing of ceramic coatings, creating databases of decorative patterns, creating pattern designs, pattern comparison between different cultures, tile cataloguing, and so on.

The use of sketch based interfaces is not so extended due to the problems they present at the moment, like the lack of robustness, repeatability and reliability. The current recognisers used on line in some interactive applications do not offer ideal solutions, they use to be rigid and the success ratio in the classification decreases when the number of symbols admitted by the system increases. This kind of interfaces could, for instance, support the initial stages of design, in which the traditional pen and paper are still necessary, achieving the link of these phases with the parametric model of the product. In this paper a reliable and robust agent-based architecture to support the user in the first conceptual design stages by recognising hand-sketched symbols is presented and evaluated.