Fruit Images Research Articles

Separation of the Mature Level of Papaya Callina Fruit Automatically Based on Color (RGB) uses Digital Image Processing

The process of sorting papaya fruit based on quality is one of the processes that greatly determines the quality of papaya fruit that will be sold to consumers. The process of identifying the quality of fruits using the human eye has the disadvantages of requiring more energy to sort, the level of human perception in terms of different fruit sorting, the level of human consistency in assessing fruit quality is also unstable because humans can experience fatigue. Research on fruit using image processing is the current trend, especially for fruit conditions, both qualities weight and size because this system processes faster and avoids or reduces failures that occur as a result of human nature. The process of selecting the level of fruit maturity in the process of recognition and determination and classification of post-harvest agricultural products on papaya fruit, depends on how the system is built. This study aims to build a quality recognition system for papaya fruits using Digital Image Processing technology, to analyze the level of color values (RGB), to determine the maturity level of papaya-callina fruit, so that later can be used as a reference in determining the maturity level of papaya fruit. First, the image of papaya is taken, or the acquisition uses a camera to be used as a database based on the condition of its maturity level. Second, the separation of the fruit image with the background based on the pixels, calculating the pixel value looking for the mean value, min, the max that is used later in the reference in determining the fruit maturity condition: young ripe, the half mature, mature. The results of this study provide information about pixel data in which young ripe papaya, red value does not dominate that is 7.785495, the green value becomes the highest value of 10.23922, papaya the half mature, it can be seen that the red and green composition of the pixel value is almost the same, namely 12.56288 and 12.12431, while the fully mature condition of the papaya, average red pixel value becomes more dominant when compared to green, which is 24,111901 for red and 13,70812 for green.

Integration of wavelet network and image processing for determination of total pigments in bitter orange (Citrus aurantium L.) peel during ripening

AbstractThis article presents a novel computer‐aided prediction system based on integration of fixed‐grid wavelet network (FGWN) and image processing technique for quantification of pigments (chlorophyll and carotenoid contents) in bitter orange peel. The proposed intelligent system captures the fruits image and uses image processing followed by principal component analysis (PCA) in order to extract the effective features. They consisted of 34 color features in the RGB and CIE color spaces, including some statistical parameters of color matrices (R, G, and B) or their differences and ratios as well as three CIE parameters. The first four principal components of the selected features are then fed as the inputs to the FGWN for estimating each pigment. The chlorophyll and carotenoid contents were modeled with the acceptable determination coefficient of 0.96 and 0.87, respectively. To evaluate the performance of the WN, it is compared against two common and powerful artificial intelligence techniques namely artificial neural network (ANN) and adaptive neuro‐fuzzy inference system (ANFIS). The results revealed that the WN could provide almost the same performance with a much lower number of parameters. Therefore, this study develops a novel attempt of strategy to be used for quality control of the bitter orange peel.Practical applicationsA research challenge could be developing a fast, cheap, and nondestructive system that predicts pigments contents of citrus fruits for different applications and/or specifying the necessity of postharvest artificial de‐greening to satisfy market demand. The integration of image processing and fixed‐grid wavelet network (FGWN) is a potential approach to overcome the mentioned challenge. Such an integration showed acceptable results for the determination of chlorophylls and carotenoids content in bitter oranges in the current study. In terms of applicability, the parsimonious models of the FGWN is quite desirable for online uses.

Single-Shot Convolution Neural Networks for Real-Time Fruit Detection Within the Tree.

Image/video processing for fruit detection in the tree using hard-coded feature extraction algorithms has shown high accuracy on fruit detection during recent years. While accurate, these approaches even with high-end hardware are still computationally intensive and too slow for real-time systems. This paper details the use of deep convolution neural networks architecture based on single-stage detectors. Using deep-learning techniques eliminates the need for hard-code specific features for specific fruit shapes, color and/or other attributes. This architecture takes the input image and divides into AxA grid, where A is a configurable hyper-parameter that defines the fineness of the grid. To each grid cell an image detection and localization algorithm is applied. Each of those cells is responsible to predict bounding boxes and confidence score for fruit (apple and pear in the case of this study) detected in that cell. We want this confidence score to be high if a fruit exists in a cell, otherwise to be zero, if no fruit is in the cell. More than 100 images of apple and pear trees were taken. Each tree image with approximately 50 fruits, that at the end resulted on more than 5000 images of apple and pear fruits each. Labeling images for training consisted on manually specifying the bounding boxes for fruits, where (x, y) are the center coordinates of the box and (w, h) are width and height. This architecture showed an accuracy of more than 90% fruit detection. Based on correlation between number of visible fruits, detected fruits on one frame and the real number of fruits on one tree, a model was created to accommodate this error rate. Processing speed is higher than 20 FPS which is fast enough for any grasping/harvesting robotic arm or other real-time applications.HIGHLIGHTSUsing new convolutional deep learning techniques based on single-shot detectors to detect and count fruits (apple and pear) within the tree canopy.

Development and an Application of Computer Vision System for Nondestructive Physical Characterization of Mangoes

This research reports the comparison study of manual (digital vernier caliper, DVC) method and application of computer vision system (CVS) for evaluation of size attributes (length, width and thickness) of two mango cultivars (Chausa and Dashehari). The color RGB (red, green and blue) camera was used in CVS for capturing of required profile image of fruit. The profile image was then processed and developed algorithms for image analysis and extraction of fruit’s quality attributes. Various morphological features such as area, perimeter, Max Feret diameter (MFD), hydraulic radius, Waddel disk diameter (WDD), elongation factor, compactness factor, Heywood circulatory factor and type factor of both cultivars were also evaluated using developed algorithms and attempted to describe the morphology of the outline and shape of the biological object. Area, perimeter, Max Feret diameter and Waddel disk diameter, further, correlated with the weight and volume of fruit. Various multilinear regression (MLR) models were developed for determining weight and volume of mango fruit of both cultivars using dimensional characteristics (length, width and thickness) of mango fruit evaluated by CVS. The major diameter (single variable) and all three (major, intermediate and minor) diameter (multiple variables)-based models were identified as the best models. Similarly, CVS has yielded above 97.9%, 93.5% and 92.5% accuracy in estimating mango fruits major, intermediate and minor diameters of both cultivars.

Improvement of Convolutional Neural Network Accuracy on Salak Classification Based Quality on Digital Image

Salak is a seasonal fruit that has high export value. The success of salak fruit exported is influence by selection process, but there is still a problem in it. The selection of salak still done manually and potentially misclassified. Research to automate the selection of salak fruit has been done before. The process of selection this salak fruits used convolutional neural network (CNN) based on image of salak fruits. The resulting of accuracy value from previous research is 70.7% for four class classification model and 81.45% for two class classification model. This research was conducted to increase accuracy value the classification of salak exported based on previous research. Accuracy improvement by changing the noise removal process to produce a better image. The changing also occur in the CNN architecture that layer convolution is more deep and with additional parameters such as Stride, Zero Padding, and Adam Optimizer. This change hopefully can increase the accuracy value of the salak classification. The results showed an accuracy value increased 22.72% from 70.70% to 93.42% for the category of four classes CNN models and increased 13,29% from 81.45% to 94.74% for category two classes.

Use of hyperspectral transmittance imaging to evaluate the internal quality of nectarines

The internal quality of nectarines (Prunus persica L. Batsch var. nucipersica) cv. ‘Big Top’ (yellow flesh) and ‘Magique’ (white flesh) has been inspected using hyperspectral transmittance imaging. Hyperspectral images of intact fruits were acquired in the spectral range of 630–900 nm using transmittance mode during their ripening under controlled conditions. The detection of split pit disorder and classification according to an established firmness threshold were performed using PLS-DA. The prediction of the Internal Quality Index (IQI) related to ripeness was performed using PLS-R. The most important variables were selected using interval-PLS. As a result, an accuracy of 94.7% was obtained in the detection of fruits with split pit of the ‘Big Top’ cultivar. Accuracies of 95.7% and 94.6% were achieved in the classification of the ‘Big Top’ and ‘Magique’ cultivars, respectively, according to the firmness threshold. The internal quality was predicted through the IQI with R2 values of 0.88 and 0.86 for the two cultivars. The results obtained indicate the great potential of hyperspectral transmittance imaging for the assessment of the internal quality of intact nectarines.

Image-based deep learning automated sorting of date fruit

Deep Convolutional Neural Network (CNN) with a unique structure for combining the feature extraction and classification stages, has been considered to be a state-of-the-art computer vision technique for classification tasks. This study presents a novel and accurate method for discriminating healthy date fruit (cv. Shahani), from defective ones. Furthermore, owing to the use of deep CNN, this method is able to predict the ripening stage of the healthy dates. The proposed CNN model was constructed from VGG-16 architecture which was followed by max-pooling, dropout, batch normalization, and dense layers. This model was trained and tested on an image dataset containing four classes, namely Khalal, Rutab, Tamar, and defective date. This dataset was collected by a smartphone under uncontrolled conditions with respect to illumination and camera parameters such as focus and camera stabilization. The CNN model was able to achieve an overall classification accuracy of 96.98%. The experimental results on the suggested model demonstrated that the CNN model outperforms the traditional classification methods that rely on feature engineering for discrimination of date fruit images.

Fractional fuzzy 2DLDA approach for pomegranate fruit grade classification

In this research work, the fractional fuzzy two-dimensional linear discriminant analysis (FF2DLDA) an extension of fuzzy 2DLDA (F2DLDA) is proposed and applied to grade the two dimensional pomegranate fruit images. The research work investigates three existing and one novel mathematical feature extraction simulation technique to address the problem of non-destructive pomegranate fruit grading and classification. The approach is to use four feature extraction based mathematical simulation techniques, includes traditional 2DLDA, Fractional 2DLDA (FLDA), Fuzzy 2DLDA and FF2DLDA. The proposed technique holds the most discriminative features, by redefining the fuzzy between class scatter matrix of F2DLDA as fractional fuzzy between-class scatter matrix. In order to classify the extracted features, the kernel support vector machine (KSVM) along with all 2DLDA variants are used. The results show that FF2DLDA is manifold superior to the existing techniques, since the fractional fuzzy between class scatter matrix assigns a small fuzzy weight for edge classes and large fuzzy weight for non-edge classes. This effectively weakens the effect of edge class selection problem which is present in traditional 2D feature extraction techniques.

Maturity detection and volume estimation of apricot using image processing technique

Physical and imaging properties of apricot fruits are the main factors considered in the design and development of sorting mechanisms. Classification of apricots based on visual appearance was performed using image processing technique. The apricots were classified into three maturity stages (i.e. unripe, ripe, and overripe) and the volume was estimated. The captured images of fruits were processed using a previously developed automatic algorithm. The images were cropped, filtered, and segmented upon which imaging features of apricots including relative R, G, B channels, gray-scale, L*, a*, and b* were extracted. The volumes of apricots were estimated using the stripping method and multiplying the value by an oval factor. The result of statistical analysis indicated that there was significant difference among the maturity stages with respect to G, gray-scale, L* and b* features. The LDA and QDA classifiers could categorize the apricots with the accuracy of 0.904 and 0.923, respectively based on color features. Results showed that the algorithm can properly classify the fruits using the image properties of apricots.

Relative Impact of Nutritional Warnings and Other Label Features on Cereal Bar Healthfulness Evaluations

To investigate the relative influence of nutritional warnings and 2 marketing strategies commonly used in food labels, nutrient claims, and fruit images on consumers' healthfulness judgments. Labels of cereal bars were designed according to a full factorial design with 3 2-level variables: images of fruit (with vs without), nutrient claims about the fiber content (with vs without), and nutritional warnings about excessive content of sugar and saturated fat (with vs without). Experiment conducted in Montevideo, Uruguay. One hundred Uruguayan people, 75% female, aged 18-56 years. Healthfulness perception and eye-tracking variables. The researchers used ANOVA to evaluate the influence of the experimental design's variables on perceived healthfulness and eye-tracking measures. Nutritional warnings caught participants' attention and reduced the amount of visual attention needed to evaluate healthfulness. Participants mainly relied on nutritional warnings to make healthfulness judgments. Findings of the current work confirm the potential of nutritional warnings to influence consumers' healthfulness perception, overriding the effect of other label cues used by the food industry to convey the concept of healthfulness.

Automatic Classification for Fruits’ Types and Identification of Rotten Ones Using k-NN and SVM

Having a system that classifies different types of fruits and identifies the quality of fruits will be of a value in various areas especially in an area of mass production of fruits’ products. This paper presents a novel system that differentiates between four fruits types and identifies the decayed ones from the fresh. The algorithms used are based on the colour and the texture features of the fruits’ images. The algorithms extract the RGB values and the first statistical order and second statistical of the Gray Level Co-occurrence Matrix (GLCM) values. To segregate between the fruits’ types, Fine, Medium, Coarse, Cosine, Cubic, and Weighted K-Nearest Neighbors algorithms are applied. The accuracy percentages of each are 96.3%, 93.8%, 25%, 83.8%, 90%, and 95% respectively. These steps are tested with 46 pictures taken from a mobile phone of seasonal fruits at the time i.e., banana, apple, and strawberry. All types were accurately identifying. To tell apart the decayed fruits from the fresh, the linear and quadratic Support Vector Machine (SVM) algorithms differentiated between them based on the colour segmentation and the texture feature algorithms values of each fruit image. The accuracy of the linear SVM is 96% and quadratic SVM 98%.

Development of a computer vision system to estimate the colour indices of Kinnow mandarins.

Maturity of a citrus fruit is generally expressed by a numerical value called citrus colour index (CCI). The success of methods employed in estimating the maturity depends on the cultivar and climatic conditions of growing regions. In this work, an image processing based method using CIELAB color model has been developed to estimate the CCI of Kinnow mandarin fruits. A polynomial transformation based camera characterization method was employed to reduce the number of transformations required for RGB to colour space transformation, which resulted into a colour difference of 2.191 with CIELAB 2000 colour difference formula. In order to analyse the performance of this method, linear regression and partial least square (PLS) models were built on a dataset of 271 Kinnow fruit images wherein spectrophotometer was used for the validation of computed CCI values. The proposed method achieved a high adjusted value of 0.9660 using PLS regression, which ascertain the feasibility of image processing based system in estimating the maturity of Kinnow fruits. Additionally, a correlation analysis between colour coordinates and physicochemical properties was conducted to analyze the relation between the fruit's external peel colour and its internal characteristics.

Automatic Fruit Classification Using Deep Learning for Industrial Applications

Fruit classification is an important task in many industrial applications. A fruit classification system may be used to help a supermarket cashier identify the fruit species and prices. It may also be used to help people decide whether specific fruit species meet their dietary requirements. In this paper, we propose an efficient framework for fruit classification using deep learning. More specifically, the framework is based on two different deep learning architectures. The first is a proposed light model of six convolutional neural network layers, whereas the second is a fine-tuned visual geometry group-16 pretrained deep learning model. Two color image datasets, one of which is publicly available, are used to evaluate the proposed framework. The first dataset (dataset 1) consists of clear fruit images, whereas the second dataset (dataset 2) contains fruit images that are challenging to classify. Classification accuracies of 99.49% and 99.75% were achieved on dataset 1 for the first and second models, respectively. On dataset 2, the first and second models obtained accuracies of 85.43% and 96.75%, respectively.

Plant size estimation based on the construction of high-density corresponding points using image registration

This paper presents an approach to estimating the plant or fruit size from objects on the digital images of natural scenes. Two images are taken from camera positions with a known distance between them. For a selected object and a pair of its boundary pixels in the first image, a corresponding pixel pair is found in the second image. The Euclidean distances within the two pairs compare in similar triangles and combine with the known distance between the camera positions to estimate the size of the object in metric units. A dense correspondence among images is first required. It can be determined by joint information of multiple 2D views of the same scene utilizing image registration techniques. These are used to align corresponding image points, without having to rely on distinct features, such as geometric properties. This is most advantageous for weak textured areas, e.g. uniform fruit shades, where many methods for corresponding point detection fail. The pixel neighbourhoods of candidate corresponding pairs are compared by template matching to verify their similarity and to select the most reliable correspondences. A stratified rigid-to-elastic registration approach generates a deformation matrix whose elements define translational vectors on a pixel basis. The accuracy of correspondence, checked by colour template matching, is assessed by mean absolute error, D, between corresponding pixel-pair intensities within the templates. By matching 26 random fruit image pairs, the proposed approach detected 1390.5 ± 1129.8 reliable corresponding points on the surface of the fruit with the accuracy of D≤5, on average. This means approximately 18% of all objects’ pixels and considerably exceeds the results of comparable methods for high-density correspondence detection, such as correlation-based correspondence, non-rigid dense correspondence (NRDC), and scale invariant feature transform (SIFT). The number and quality of corresponding points obtained by the proposed algorithm turned out to be reliable and sufficiently robust for an accurate estimation of distances between objects and camera positions (with an overall accuracy of 0.11 m ± 0.06 m) and height of plants (with an accuracy of 0.14 m ± 0.1 m) when taking two similar outdoor photographs of a scene.

Identification of Artificially Ripened Fruits Using Machine Learning

Ripening of fruit is a natural process. Ethylene is responsible for ripening process which is produced naturally in fruit. But dealers and sellers often use chemicals like CaC2 (Calcium carbide) to quicken this process so their product will enter the market early and they can maximize profit. Fruits are kept in storage with chemicals. This chemical mixes with moisture and produces ethylene which causes ripening of the fruit. Since the ethylene is high in quantity and it contacts with the surface area of the fruit it causes uniform ripening of the fruit unlike when the fruit is ripened naturally it causes uneven ripening of the fruit since natural ethylene present in the fruits is un-uniformly distributed. Consuming such fruit is harmful to human health. It can cause headache, stomach irritation, throat irritation, digestion problem and since the chemical used is carcinogenic it can also cause cancer and it also degrades the taste and quality of the fruit. Finding out the artificially ripened fruit with the human eye is difficult. So, the proposed system gets an image of fruit under the test and compare it with the features of naturally ripened fruit and artificially ripened fruit and give the output with the probability. This method makes usage of smartphone which runs the android application and the convolutional neural network to detect the artificially ripened fruit.

Machine vision for characterisation of some phenomic features of plant parts in distinguishing varieties - a review

Phenomic features of plant parts are important varietal traits for all crops and form part of the distinctiveness, uniformity and stability (DUS) characterisation protocols. Manual methods for measurement of these traits are expensive, less consistent and time consuming hence machine vision has been used in recent researches. The machine vision systems employed for this purpose consist of acquisition systems (hardware) and image processing and analysis system (software). The area of machine vision has developed during last few decades and there have been many improvements in the employed hardware and software. Major work has been reported on use of seed and fruit images; however, images of other plant parts like leaves, roots, flowers etc. have also been used. A variety of techniques have been reported for analysis of features in order to distinguish among different crop varieties.

Machine vision for characterisation of some phenomic features of plant parts in distinguishing varieties - a review

Phenomic features of plant parts are important varietal traits for all crops and form part of the distinctiveness, uniformity and stability (DUS) characterisation protocols. Manual methods for measurement of these traits are expensive, less consistent and time consuming hence machine vision has been used in recent researches. The machine vision systems employed for this purpose consist of acquisition systems (hardware) and image processing and analysis system (software). The area of machine vision has developed during last few decades and there have been many improvements in the employed hardware and software. Major work has been reported on use of seed and fruit images; however, images of other plant parts like leaves, roots, flowers etc. have also been used. A variety of techniques have been reported for analysis of features in order to distinguish among different crop varieties.

Automatic Inspection of the External Quality of the Date Fruit

Assessing the quality of date fruits manually is a labour-intensive task. Furthermore, quality of the date fruits stored in the storage may degrade with time and it is important to inspect the quality of date routinely. After date harvesting, the industrial companies start the inspection of dates. Inspection of date fruit is to isolate the damaged or defected dates from the good one. Industrial factories demand a high quality and fast production that can only be achieved by automatic inspection. In this paper, we propose a framework for inspection of the external quality of the date fruits through the image processing method. Images of the date fruit are classified into good quality and sugar-defected by using bag of feature (BOF) method. The methodology includes five main stages including key points detection, feature extraction, creating a dictionary, vector quantization and classification. Proposed framework is tested on the dataset collected by the authors using two types of key-points detection methods. Best classification accuracy for grid based search was 99% and 84% for SURF based key-points detection method on the testing dataset.

Olive-Fruit Variety Classification by Means of Image Processing and Convolutional Neural Networks

The automation of classification and grading of horticultural products attending to different features comprises a major challenge in food industry. Thus, focused on the olive sector, which boasts of a huge range of cultivars, it is proposed a methodology for olive-fruit variety classification, approaching it as an image classification problem. To that purpose, 2,800 fruits belonging to seven different olive varieties were photographed. After processing these initial captures by means of image processing techniques, the resulting set of images of individual fruits were used to train, and continuedly to externally validate, the implementations of six different Convolutional Neural Networks architectures. This, in order to compute the classifiers with which perform the variety categorization of the fruits. Remarkable hit rates were obtained after testing the classifiers on the corresponding external validation sets. Thus, it was yielded a top accuracy of 95.91% when using the Inception-ResnetV2 architecture. The results suggest that the proposed methodology, once integrated into industrial conveyor belts, promises to be an advanced solution to postharvest olive-fruit processing and classification.

Real Life Implementation of Object Detection and Classification Using Deep Learning and Robotic Arm

Deep learning is one of most favourable domain in today’s era of computer science. In this paper we discussed, the implementation of deep learning concepts by using Auduino uno with robotic application. For the purpose of object detection and classification, a robotic arm is used in the project which is controlled to automatically detect and classify of different object (fruits in our project). In this project, the camera will capture an image of fruit for further processing in the model based on convolutional neural network (CNN). When the trained model will detect the object in image, a particular signal will be sent to robotic arm using Arduino uno, which will place the detected object into a basket. In this way our project will recognize and classify two different fruits and will place it into different baskets. This project is a demonstration of combination of deep learning concept together with Arduino programming, which itself is a complete framework. This combination can be used to solve so many real life problems. After implementation, we found up to 99.22% of accuracy in object detection.