Search

Maturity of the coconut is judged manually based on the shape, color, weight, shaking sound, timeframe, etc. Currently, there is not much research being done to find solutions involving the latest techniques including fuzzy, machine learning and deep learning techniques for identifying maturity of the coconuts without human intervention. These techniques have significant challenges in detecting the multiple classes in coconut maturity due to difference in size and shape of the coconuts and the complex nature of their backgrounds as each coconut tree and each coconut bunch are unique. In this research work, we propose an integrated fuzzy and deep learning model (IFDM) to identify the maturity classes of coconut. Deep learning-based Mask R-CNN is used to classify and segment the coconut images. The combined fuzzy system (CFS) extracts the color, shape and scratch features with a fuzzy extraction technique for each feature. A probability-based fuzzy integration (PFI) method integrates the output from each fuzzy extraction to classify the coconuts. Decision-making probability (DMP) model combines the class probability of the probability-based fuzzy integration method and Mask R-CNN model to obtain the final coconut maturity class. The dataset with 10,000 images is used for training. Two test datasets, the first one containing 1754 real-time coconut images and the second containing 1000 cropped coconut images from the background, are used for testing. Using the proposed learning model in real time, an accuracy of 86.3% is obtained in classifying the maturity classes of coconuts. While integrating the models like Yolo v3, Yolo v5, Faster R-CNN, random forest and KNN with combined fuzzy system, Mask R-CNN performed better in terms of accuracy. [ABSTRACT FROM AUTHOR]

Mango is an imperative commercial fruit in terms of market value and volume of production. In addition, it is grown in more than ninety nations around the globe. Consequently, the demand for effective grading and sorting has increased, ever since. This communication describes a non-invasive mango fruit grading and sorting model that utilizes hybrid soft computing approach. Artificial neural networks (ANN), optimized with Antlion optimizer (ALO), are used as a classification tool. The quality of mangoes is evaluated according to four grading parameters: size (volume and morphology), maturity (ripe/unripe), defect (defective/healthy) and variety (cultivar). Besides, a comparison of proposed grading system with state-of-the-art models is performed. The system showed an overall classification rate of 95.8% and outperformed the other models. Results demonstrate the effectiveness of proposed model in fruit grading and sorting applications. [ABSTRACT FROM AUTHOR]

Tomato is an important vegetable crop which provides approximately 40% of the daily recommended vitamin C and vitamin A, which supports human health. The experiments were laid out in a Completely Randomized Design (CRD) having four treatments and four time repetition under high tunnel. Nitrogen (N), phosphorus (P) and potassium (K) fertilizers were applied at the rate of 60:60:60 kg/ha or 1:1:1 to T1, T2 and T3 after each 10, 20 and 30 day's interval while Zn (5 kg/ha) was applied as a basal dose. The highest marketable yield was observed for T2 after each 20 days of NPK application. Maximum Grad-A fruit among the treatments was recorded for T1, Grade-B for T2 & T3 and Grade-C for T0. Highest density was recorded for T2, % moisture for T0, maximum chlorophyll content in leaves was recorded for T2 while highest shelf life was noted for T1, T2 and T3 except T0. The fruit and straw of tomato was analyzed to determine N, P and K (%) by using standard protocol. Among all the treatments highest N, P and K in fruit was observed for T1. In case of straw maximum N was observed for T1 whereas high level of P and K was recorded for T2. Similarly the total NPK uptake were in the order as: T1 > T2 > T3 >T0, respectively. Experiential findings revealed that application of NPK after each 20 days are the suitable interval to get high marketable yield and highly bio-fortified and quality tomato crop. [ABSTRACT FROM AUTHOR]

Post-harvest grading is an essential and important process that affects the fruit quality, evaluation, health-intensive, and export market. Even though the sorting and grading can be performed by the human as going on, it is tedious, labor-intensive, slow, and error-prone. Hence, smart automation is required for the same. Computer vision advancements touching every area where there even a minimal chance of smart automation. In this paper, intelligent automation for mango fruit grading designed and developed. Initially, the fruit segmentation is done by the active contour model, and the abnormality segmentation is performed using enhanced fuzzy-based K-means clustering approach followed by features: discrete Fourier transform (DFT), local binary pattern (LBP), and gray-level co-occurrence matrix (GLCM) and shape features extraction. The self-adaptive chicken swarm optimization (SA-CSO) has been used to reduce and optimize features vector. The quality of the fruits has finally categorized based on surface defect and maturity classification. Hence, for defect classifications, optimal abnormality segmented features have been fed to the K-nearest neighbors (KNN). The optimally selected fruit segmented features are subjected to a fuzzy classifier and the fruit segmented images are subjected to a convolutional neural network (CNN). As an improvement, the proposed SA-CSO is used to optimize the hybrid classifier for maximizing the accuracy of classification. The maturity is classified using the hybrid fuzzy classifier and CNN as ripe, partially ripe, and unripe. Finally, the defect and maturity output have been used to decide the quality as good, average, and bad. The comparative analysis of diverse performance metrics proves the effectiveness of the proposed model over other traditional algorithms. [ABSTRACT FROM AUTHOR]

In horticulture, measuring, sorting by shape, and determining the size and volume of fruits are all essential processes for meeting market quality standards and increasing market value. Fruit sorting and grading processes are very laborious and time-consuming task but machine vision-based fruit grading systems have the potential to replace human labour. However, a great challenge in vision-based fruit grading system is the recognition of different features such as shape, size, skin flaw and sometimes even three-dimensional (3D) shape. In this study, a simple and efficient image processing algorithm is proposed for estimating volume and 3D shape of mango fruit. The width and length of mango fruit are obtained from two-dimensional (2D) colour image. Then, fruit thickness is estimated based on light intensity distribution in 2D (top view) of mango fruit and maximum width–thickness correlation. The 3D shape of the mango fruit is then reconstructed. The accuracy of proposed method was compare with two existing volume prediction methods. Estimated volumes were compared with measured measurements using water-displacement method and the reconstructed 3D shapes were compared with measured structures of mango fruits. For a total of 150 mango samples, the results show that the proposed method gave an accuracy of 96.8% whilst the two other methods gave 91.7% and 91.5% respectively. The reconstructed mango shapes were therefore in close agreement with measured shapes. Image 1 • Thickness of mango in 2D image is estimated based on light intensity distribution. • Reconstructed 3D shapes of mangoes in close agreement with measured shapes. • Proposed volume estimation method gives an average accuracy of 96.8%. • Proposed method has accuracy of 93.2% for mass estimation with fixed density. [ABSTRACT FROM AUTHOR]

Grading and quality assessment is an important aspect of post-harvest management in pomegranate fruits. In India, the quality assessment is usually performed manually by feeling the fruit in hand. This manual assessment poses a lot of disadvantages such as uncertainty, tediousness, time consumption etc. Moreover there are no well-organized grading systems for testing quality of pomegranates. Aim of the present research work is to eliminate such problems associated with manual quality assessment by incorporating Machine Intelligence and Digital Image Processing techniques. The present work precisely redefines the new quality parameters associated with the existing grading criteria. The research work also proposes a unique Effective Quality Assessment (EQA) algorithm comprising of a holistic approach towards the grading and quality assessment of pomegranate fruits. Results of the research work are found to be 97.83% by using Artificial Neural Networks. [ABSTRACT FROM AUTHOR]

This paper presents a new methodology for the estimation of olive-fruit mass and size, characterized by its major and minor axis length, by using image analysis techniques. First, different sets of olives from the varieties Picual and Arbequina were photographed in the laboratory. An original algorithm based on mathematical morphology and statistical thresholding was developed for segmenting the acquired images. The estimation models for the three targeted features, specifically for each variety, were established by linearly correlating the information extracted from the segmentations to objective reference measurement. The performance of the models was evaluated on external validation sets, giving relative errors of 0.86% for the major axis, 0.09% for the minor axis and 0.78% for mass in the case of the Arbequina variety; analogously, relative errors of 0.03%, 0.29% and 2.39% were annotated for Picual. Additionally, global feature estimation models, applicable to both varieties, were also tried, providing comparable or even better performance than the variety-specific ones. Attending to the achieved accuracy, it can be concluded that the proposed method represents a first step in the development of a low-cost, automated and non-invasive system for olive-fruit characterization in industrial processing chains. [ABSTRACT FROM AUTHOR]

Abstract: Background and Aims: Evaluation of dried grape berry quality in Australia is currently based on visual assessment by trained humans, a process that is subjective and prone to error. The goal of this work was to develop an objective methodology to evaluate the quality of dried grapes from Vitis vinifera L. cv. Thompson Seedless, referred to as ‘sultanas’. Methods and Results: A non‐destructive method for berry quality assessment based on digital image analysis using the MATLAB programing language enabled the development of a novel quality index and computer application for classification of berry quality based on surface colour. The method separated individual berries into several colour classes and assigned the batch an overall crown grade that was consistent with industry assessment. Conclusion: Digital image analysis of berry surface colour using the new method was an accurate and reliable method for objectively evaluating dried sultana quality. Significance of the Study: This method will facilitate the rapid and objective evaluation of sultana quality at packinghouses to ensure that a fair price is paid to growers and to reduce potential conflicts arising from subjective fruit grading by humans. [ABSTRACT FROM AUTHOR]

Grading of fruits is the post harvesting step involving the labeling of fruits into their respective categories using external (color, size, shape etc) and internal features(aroma, taste, pH etc). This could be done manually which is tiresome, inconsistent and time consuming task and therefore computer vision based fruit grading comes into play which makes the whole grading process consistent, labor saving and fast as well. In past years, researchers had introduced many non destructive image processing technique to grade the food products, ensuring quality of food products. This paper presents algorithm of classifying guava fruit into categories using Local Tetra patterns. Features extracted are classified using SVM, K-NN, Trees, ANNs. Results generated by different classifiers are analyzed using accuracy and error rate. Linear layer neural network gave the highest results amongst all the classifiers. [ABSTRACT FROM AUTHOR]

Papaya grading is performed manually which may lead to misclassifications, resulting in fruit boxes with different maturity stages. The objective is to predict the ripening of the papaya fruit using digital imaging and random forests. A series of physical/chemical analyses are carried out and true maturity stage is derived from pulp firmness measurements. Imaging and image analysis provides hand-crafted color features computed from the peel and random decision forests are implemented to predict ripening stage. More specifically, a total of 114 samples from 57 fruits are used for the experiments, and classified into three stages of maturity. After image acquisition and analysis, twenty-one hand-crafted color features (comprising seven groups) that have low computational cost are extracted and evaluated. Random forests with two datasets (cross-validation and prediction set) are employed for the experiments. Concerning all image features, 94.3% classification performance is obtained over the cross-validation set. The prediction set obtained 94.7% misclassifying only a single sample. For the group comparisons, the normalized mean of the RGB (red, green, blue) color space achieved better performance (78.1%). Essentially, the technique can mature into an industrial application with the right integration framework. [ABSTRACT FROM AUTHOR]

To meet the demand of fruit farmers′ timely grading of navel oranges, a set of navel orange postharvest field grading system was designed based on machine vision, which was composed of conveying system, visual system and sorting system. The size of navel orange, the number of surface defects and the defect area were detected by the proposed system. The fruit was automatically graded according to the predetermined comprehensive assessment standard. The experimental results show that the average detection time of single picture is less than 30 ms, and the detection error of size is less than 3% with defect detection rate of 99% and defect area detection error less than 7%. Compared with the traditional detection system, the proposed system has the advantages of high detection speed and compact design, which is suitable for in-filed grading of navel orange. [ABSTRACT FROM AUTHOR]

With increasing demand of fruits of great quality at per safety standards, necessity for valid, speedy, unbiased class calculation of these characteristics in fruits pursued to arise. For producers, Fruit grading is an vital operation since it influence the export market and fruits quality assessment. It involves qualitative analysis of fruit considering both internal and external factors where internal factors involves the nutritive values, sweetness, aroma, flavor etc and external factors examine the fruit from visual inspection using shape, texture, size, color etc. Manual sorting and grading of fruits is slow, tiresome, error prone and needing large workforce, thus there is a requirement of a smart, self regulating fruit grading system. Researchers have unfolded various algorithms in latter years. This paper presents the relevant components of a computer vision system and emphasizes the important aspects of the image processing technique used in qualitative inspection of fruits by examining the external features of different fruits. [ABSTRACT FROM AUTHOR]

The relationship between berry chemical composition, region of origin and quality grade was investigated for Chardonnay grapes sourced from vineyards located in seven South Australian Geographical Indications (GI). Measurements of basic chemical parameters, amino acids, elements, and free and bound volatiles were conducted for grapes collected during 2015 and 2016. Multiple factor analysis (MFA) was used to determine the sets of data that best discriminated each GI and quality grade. Important components for the discrimination of grapes based on GI were 2-phenylethanol, benzyl alcohol and C6 compounds, as well as Cu, Zn, and Mg, titratable acidity (TA), total soluble solids (TSS), and pH. Discriminant analysis (DA) based on MFA results correctly classified 100% of the samples into GI in 2015 and 2016. Classification according to grade was achieved based on the results for elements such as Cu, Na, Fe, volatiles including C6 and aryl alcohols, hydrolytically-released volatiles such as (Z)-linalool oxide and vitispirane, pH, TSS, alanine and proline. Correct classification through DA according to grade was 100% for both vintages. Significant correlations were observed between climate, GI, grade, and berry composition. Climate influenced the synthesis of free and bound volatiles as well as amino acids, sugars, and acids, as a result of higher temperatures and precipitation. [ABSTRACT FROM AUTHOR]

Determination of optimum maturity and ripeness of fruit is essential in the production of processed fruit, including pineapples, but this is difficult to achieve consistently by visual grading in commercial factories. Therefore, this study tested two nondestructive techniques for predicting the maturity index of intact pineapple. These were transmittance short wavelength near infrared spectroscopy (SW-NIRS) in the wavelength range of 665–955 nm and reflectance near infrared hyperspectral imaging (NIR-HSI) in the wavelength range of 935–1720 nm. The number of samples used for calibration was 120 for both SW-NIRS and NIR-HSI. The maturity index and spectral information of individual pineapple fruit were acquired from both techniques and analysed using the same procedure. Then, partial least squares regression (PLSR) was used to establish the models for predicting the maturity index of each intact fruit. The leave-one-out cross validation was used for evaluating the performance of the models. The results showed that both techniques gave reliable performance in predicting the maturity index of individual fruit, with a coefficient of determination considering cross validation (R cv 2) for the prediction of the maturity index of 0.70 and a root mean square error in cross validation (RMSECV) of 2.16 when using SW-NIRS and R cv 2 of 0.72 and RMSECV of 1.68 when using NIR-HSI. It was therefore concluded that both SW-NIRS and NIR-HSI had the potential for use in nondestructive analysis of the maturity of intact pineapple fruit in fruit processing factories. [Display omitted] • SW-NIRS and NIR-HSI were used for evaluating pineapple quality. • SW-NIRS and NIR-HSI could predict TSS/TA (R cv ² = 0.70 and 0.72). [ABSTRACT FROM AUTHOR]

This study aims to demonstrate the feasibility for classifying kiwifruit into shape grades by adding a single camera to current Chinese sorting lines equipped with weight sensors. Image processing methods are employed to calculate fruit length, maximum diameter of the equatorial section, and projected area. A stepwise multiple linear regression method is applied to select significant variables for predicting minimum diameter of the equatorial section and volume and to establish corresponding estimation models. Results show that length, maximum diameter of the equatorial section and weight are selected to predict the minimum diameter of the equatorial section, with the coefficient of determination of only 0.82 when compared to manual measurements. Weight and length are then selected to estimate the volume, which is in good agreement with the measured one with the coefficient of determination of 0.98. Fruit classification based on the estimated minimum diameter of the equatorial section achieves a low success rate of 84.6%, which is significantly improved using a linear combination of the length/maximum diameter of the equatorial section and projected area/length ratios, reaching 98.3%. Thus, it is possible for Chinese kiwifruit sorting lines to reach international standards of grading kiwifruit on fruit shape classification by adding a single camera. [ABSTRACT FROM AUTHOR]

It is said that the backbone of Indian economy is agriculture. The contribution of the agriculture sector to the national GDP (Gross Domestic Products) was 14.6% in the year 2010. To attain a growth rate equivalent to that of industry (viz., about 9%), it is highly mandatory for Indian agriculture to modernize and use automation at various stages of cultivation and post-harvesting techniques. The use of computers in assessing the quality of fruits is one of the major activities in post-harvesting technology. As of now, this assessment is majorly done manually, except for a few fruits. Currently, the fruit quality assessment by machine vision in India is still at research level. Major research has been carried out in countries like China, Malaysia, UK, and Netherlands. To suit the Indian market and psychology of Indian farmers, it is necessary to develop indigenous technology. This paper is the first step toward evaluating the research carried out by the research community all over world for tropical fruits. For the purpose of survey, we have concentrated on the tropical fruits of the state of Maharashtra, while keeping in focus of the review image processing algorithms. [ABSTRACT FROM AUTHOR]

Kiwi fruit grading is a key link of fruit treatment after picking. In respect that low sorting rate and grading rate of kiwi fruits affect commodity value in China, the paper acquires the volumetric characteristic of kiwi fruit through the 3D reconstruction of its three views, and fits volumetric pixel P with actual weight W to get a mathematical model between the two parameters, and identifies the area characteristic of the surface defect of kiwi fruit by iterative method. Volumetric characteristic and the characteristic of the surface defect are inputted into Support Vector Machine (SVM), and then suitable kernel function and kernel parameters are selected, finally the kiwi fruit grading accuracy of 97.73% is gotten. Compared with the grading accuracy of BP network with 92.11%, it can be seen that the grading effect of SVM is better than that of BP network and its grading performance is stable. The experimental results show that the SVM grading method based on kiwi fruit volume and its surface defect is feasible and can be used for online inspection of appearance quality of kiwi fruits. [ABSTRACT FROM AUTHOR]

Highlights: [•] Vis/SWNIR spectroscopy and spectral scattering were used for sorting three apple varieties into two quality grades. [•] A sensitivity analysis was performed using different quality threshold criteria for apple sorting. [•] Good sorting results for firmness and moderate for SSC were obtained using an online scattering technique. [•] Better sorting results for firmness and SSC were obtained using a stationary Vis/SWNIR technique. [•] When the classification involves firmness and SSC quality attributes, the sorting error increases. [Copyright &y& Elsevier]

Wearable computing is a form of ubiquitous computing that offers flexible and useful tools for users. Specifically, glove-based systems have been used in the last 30 years in a variety of applications, but mostly focusing on sensing people's attributes, such as finger bending and heart rate. In contrast, we propose in this work a novel flexible and reconfigurable instrumentation platform in the form of a glove, which can be used to analyze and measure attributes of fruits by just pointing or touching them with the proposed glove. An architecture for such a platform is designed and its application for intuitive fruit grading is also presented, including experimental results for several fruits. [ABSTRACT FROM AUTHOR]

Abstract: 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. [ABSTRACT FROM AUTHOR]

Abstract: The maturity of nectarines at harvest can be assessed by measuring the absorption coefficient at 670nm (μ a) with the non-destructive technique of time-resolved reflectance spectroscopy (TRS). A kinetic model links μ a, converted into the biological shift factor (BSF), to firmness decrease during ripening; in this way the firmness decay model includes the variations in maturity at harvest, thereby allowing prediction of shelf-life for individual fruit. In order to study how this methodology could be practically used at the time of harvest, when μ a can be measured non-destructively on all fruit, while the destructive measurement of firmness can only be done on a small sample, various firmness decay models were developed using either data at harvest or within 1–2 d after harvest from previous experimental research with nectarines carried out over a 5-year period. These models were then tested for prediction and classification ability by comparing the predicted firmness and class of usability to the actual ones measured during ripening and their performance compared to that of models based on data during the whole shelf-life. Our results suggest that the methodology might be used as a management tool in the nectarine supply chain. Independently from the actual softening rate, the classification at harvest based on μ a is able to segregate fruit of different quality and maturity according to their softening behaviour during shelf-life. Among the various models, those estimated using data at harvest and after 24h of shelf-life had better performance than those based only on data at harvest. In the 2002 and 2005 seasons, this model showed a classification ability very close to that of models based on data during the whole shelf-life. However, its performance in the 2004 season was not so good, because it could not take into account the influence of cold storage periods prior to shelf-life. All the steps necessary to apply this methodology are detailed. [Copyright &y& Elsevier]

The terminal velocity of kiwifruit falling through a fluid is largely dependent on the density and hence the dry matter content of the fruit, but is also almost independent of fruit size and shape. This article shows how a water-based flume could be used to separate kiwifruit for its internal (taste-related) properties in bulk quantities and at high speed, obviating the need to use density-enhancing salts, which are less desirable. The mathematics of flume operation are discussed, and trials of a laboratory flume were conducted. It is shown that below a density of 1020 kg m-3 (dry matter approximately 12.5% for kiwifruit), the standard error of the separation reaches 5 kg m-3 (~1% dry matter). This may be adequate to provide useful removal of low-quality kiwifruit from a line that fails grade standards. The flume sorting standard deviation increases with density, reaching 13 kg m-3 at 1045 kg m-3 (2.5% at 18% dry matter). This higher sorting error may limit the use of the technique for separating better tasting, high dry matter fruit from a product stream unless the fluid has its density artificially increased by the addition of salts, which introduces operational disadvantages. [ABSTRACT FROM AUTHOR]

Cultivar discrimination during on-line quality selection is required by high quality food industries. The aim of this work was to evaluate the potential use and efficacy of shape-based techniques in order to discriminate among four traditional Italian cultivars (Tonda di Giffoni, San Giovanni, Mortarella and Tonda Romana). Tonda di Giffoni and Tonda Romana are very similar having a spherical shape, while the other two cultivars are elongated. Color RGB images of about 400 hazelnuts were analysed with a morphological method based on the elliptic Fourier approximation to closed contours in a two-dimensional plane. This method was applied on the three outlines obtained by the polar, lateral and random plane positioning view of in-shell and unblanched kernel. The coefficients of the harmonic equations were analysed via Partial Least Square Discriminant Analysis (PLSDA) multivariate classification and mean outline for each group was graphically extracted. Results show higher percentage of correct classification for the lateral view (from 77.5% to 98.8% in the independent test). Also the random positioning view, in particular for in-shell kernels between the two rounded cultivars and between the two oblong cultivars, showed good classification results (respectively, 95.1 and 97.6). This preliminary study demonstrates the potential of modern multivariate techniques using shape-based methods on digital images to achieve high efficiency performance in fruit grading and classification. [Copyright &y& Elsevier]

Abstract: Software for detecting the quality features of golden delicious starfruits of Averrhoa carambola L. genus were developed for automated inspection system using machine vision technology. The features considered were colour and shape. The use of artificial classifiers such as linear discrimination analysis and multi-layer perceptron neural network as a tool to detect starfruit maturities such as unripe, underripe, ripe and overripe in HSI colour space were investigated. The colour spectra of matured and unmatured fruits were characterised using all colour features ranging from hue 10 to hue 74, and, using principal hues generated by Wilks-lambda analysis. Experiments performed on 200 independent starfruit samples revealed that linear discriminant analysis after Wilks-lambda analysis was more precise in classification than direct application of linear discriminant analysis. However, the classification accuracy of multi-layer preceptron remained relatively the same before and after feature reduction. Overall, the average correct classification for linear discriminant analysis and multi-layer preceptron was 95.3% and 90.5% respectively during testing stage. Meanwhile the use of Fourier transform was investigated for shape discrimination. This algorithm produced 100% success rate in detecting starfruits by three shape categories: well-formed, slightly deformed and seriously deformed. Both colour and shape analysis was easily affected by the variation of lighting levels, and this contributed to the major classification error. [Copyright &y& Elsevier]

Two fully replicated trials were conducted with glasshouse-grown tomatoes, under conditions similar to commercial production, to define the impact of Pepino mosaic virus (PepMV). PepMV was not found to reduce bulk yields in these trials, but the quality of tomato fruits harvested was reduced significantly. Compared with uninoculated, PepMV-free control plants, 6·5% of fruits of PepMV-affected cv. Espero were downgraded from class 1 in trial 1. In trial 2, an average 38% of class 1 fruits from PepMV-affected cvs Espero and Encore were lost as a result of downgrading. Loss of quality was mainly a result of blotchy ripening, gold marbling, gold spot, and symptoms directly attributed to PepMV infection. PepMV infection also affected fruit size. The results are discussed in relation to the demands of multiple retailers in the UK for class 1 tomatoes only. [ABSTRACT FROM AUTHOR]

Abstract: Brown core is an internal disorder sometimes seen in pears under controlled atmosphere (CA) storage. The symptoms are not externally recognizable and visible only after cutting the fruit. Development of nondestructive measurements of brown core of pears would benefit producers, processors, and packers. A technique using body transmittance in the 651–1282nm region was evaluated as a nondestructive technique for identifying the Chinese pear ‘Yali’ with brown core. Based on the 651–1282nm region, discriminant analysis with Mahalanobis Distance (MD) analysis can discriminate between brown core and normal pears and grade the brown core pears to three classes—slight, moderate, and severe. None of the good pears was misclassified as brown core, and none of pears with brown core was misclassified as normal pears. Exhaustive searches of the best combination of individual chrematistic wavelengths were performed on a set of 66 pears from two categories–good and brown core. The best modeling classification occurred for precisely aligned pears using the difference at two wavelengths, 713 and 743nm. When optical density (OD) difference between 713 and 743nm was applied to a test set, the classification model correctly identified the good or brown core with 95.4% accuracy. Only 5.3% of the good pears were incorrectly classified, 4.3% error of pear with brown core were classified as good fruit. [Copyright &y& Elsevier]

A prototype machine vision station was developed for quality grading of the unique star shape and rich golden colour carambola fruits (Averrhoa carambola L.). The grading criteria were based on the Malaysia's Federal Agriculture Marketing Authority standards for starfruits export. A Fourier-based shape separations method was developed for shape grading, whereas colour recognition was established using multivariate discriminant analysis. The Wilks' lambda analysis was invoked to transform and compress the data set comprising of large number of interconnected variates to a reduced set of variates. A robust scaling normalization was introduced in order to achieve invariant to geometrical transformations. Over 200 starfruit samples were inspected by the machine vision system and the results were compared to human judgement. Overall, the vision system was able to classify correctly 100% of the starfruits for shape and 92% for colour. Since the algorithms were implemented in software, the system could be programmed to inspect other fruits and agriculture commodities. [ABSTRACT FROM AUTHOR]

This paper describes an experimental study on non-destructive methods for sorting peaches according to their degree of ripeness. The method is based on near-infrared (NIR) transmittance spectrometry in the region between 730 and 900 nm. It estimates the ripeness in terms of internal sugar content and firmness. A station for acquiring the NIR signal has been designed and realized, carefully choosing between several options for each component. Four different stations have been realized and compared during the experimental phase. The signals acquired by the station have been pre-processed using a noise-reducing method based on a packets-wavelet transform. In addition, an outlier detection technique has been applied for identifying irregular behaviors inside each of the considered classes. Finally, a minimum distance classifier estimates the grade of each experimental data. The results obtained in classification show that this early version of the station enables the correct discrimination of peaches with a percentage of 82.5%. [Copyright &y& Elsevier]

This paper presents a hierarchical grading method applied to Jonagold apples. Several images covering the whole surface of the fruits were acquired thanks to a prototype grading machine. These images were then segmented and the features of the defects were extracted. During a learning procedure, the objects were classified into clusters by k-mean clustering. The classification probabilities of the objects were summarised and on this basis the fruits were graded using quadratic discriminant analysis. The fruits were correctly graded with a rate of 73%. The errors were found having origins in the segmentation of the defects or for a particular wound, in a confusion with the calyx end. [Copyright &y& Elsevier]

With increased expectations for food products of high quality and safety standards, the need for accurate, fast and objective quality determination of these characteristics in food products continues to grow. Computer vision provides one alternative for an automated, non-destructive and cost-effective technique to accomplish these requirements. This inspection approach based on image analysis and processing has found a variety of different applications in the food industry. Considerable research has highlighted its potential for the inspection and grading of fruits and vegetables. Computer vision has been successfully adopted for the quality analysis of meat and fish, pizza, cheese, and bread. Likewise grain quality and characteristics have been examined by this technique. This paper presents the significant elements of a computer vision system and emphasises the important aspects of the image processing technique coupled with a review of the most recent developments throughout the food industry. [Copyright &y& Elsevier]

Three experiments were conducted on 'Empire' apple (Malus xdomestica Borkh.) to evaluate the effects of hard water, calcium chloride (CaCl[sub2]), water conditioners, surfactants, and captan fungicide on the growth reduction and fruit cracking caused by prohexadione-calcium (PC). Two applications of 63 mg⋅L[sup-1] PC provided season-long growth control in two studies. Adding a water conditioner to PC reduced shoot growth more than an application of PC in hard or soft water in one New York study. Ammonium sulfate (AMS) and Choice were equally effective water conditioners. PC provided no growth control of water sprouts and had no effect on fruit set or yield. PC applied at 250 mg⋅L[sup-1] reduced fruit size. 'Empire' fruit cracking and corking was severe, despite the use of only 63 mg⋅L[sup-1] PC in two of the three experiments. This damage was exacerbated by the addition of a water conditioner, however AMS applied with a surfactant but without PC had little or no effect on either the severity or extent of fruit injury. In a third experiment, the addition of surfactants, CaCl[sub2], or captan to 250 mg⋅L[sup-1] PC plus a water conditioner had no effect on the severity of fruit damage. Fruit cracking caused by PC increased preharvest drop in two of three experiments, and increased postharvest rot in the Geneva, N.Y., experiment where fruit were stored prior to grading. Application of PC plus a water conditioner reduced estimated gross return per hectare for 'Empire'. We conclude that the fruit injury is caused by the formulated PC product itself under certain environmental conditions, and that this product should not be used on 'Empire'. Chemical name used: calcium 3-oxido-4-proprionyl-5-oxo-3-cyclohexine-carboxylate [prohexadione-calcium (PC)]. [ABSTRACT FROM AUTHOR]

• Accurate 3-variety automatic computer vision orange pH estimation system. • Properly validated pH values over 300 orange true measured pH and image database. • Non-intrusive computer vision assessment of pH values in orange:Bam, Blood, Thomson. • Statistical significant results: regression, error coefficients, and pH boxplots. • Potential application as grading prototype under industrial environment conditions. An automatic 3-variety computer vision orange fruit pH value assessment system in the visible-range is presented, including each 100 different color images from Bam, Blood and Thomson orange of which the true pH has been measured and is known in advance. A total of 452 features are extracted from segmented orange color images. Results with repeated trials include: true versus estimated mean pH values, true minus estimated pH values boxplots, fitness regression dispersion plots and various error measure boxplots, for both single orange variety as well as the three-orange varieties altogether (test set), showing consistent results over all three orange varieties. Regression coefficient for pH estimation in Bam, Blood and Thomson orange varieties, were 0.950, 0.935 and 0.957, respectively. Results show that the hybrid ANN-ABC estimates pH values in orange quite similarly among orange varieties, implying that properly pH values estimation is possible for different orange varieties, regardless of orange type. [ABSTRACT FROM AUTHOR]

Lycopene is one of the most important functional components produced in tomatoes. Traditional methods for determining lycopene content usually rely on destructive sampling followed by chemical analysis. This report introduces the use of a smart device camera combined with a new Android application as a novel, rapid, and nondestructive method for determining tomato lycopene content and fruit grade. We first measured chromaticity values and lycopene content in 60 tomato fruits to develop an effective predictive model for lycopene estimation. Second, a method based on color differences between test fruit and several standard colors was designed as a color classification algorithm for fruit grading. Third, an Android application for lycopene estimation and fruit grading was developed. Finally, application functionality was evaluated using an Android 4.2.2 tablet. The results demonstrated that this new approach was functional and its ability to estimate lycopene content and fruit grade was satisfactory. [ABSTRACT FROM AUTHOR]