Fruit Classification Research Articles

Utilisation of Deep Learning with Multimodal Data Fusion for Determination of Pineapple Quality Using Thermal Imaging

Fruit quality is an important aspect in determining the consumer preference in the supply chain. Thermal imaging was used to determine different pineapple varieties according to the physicochemical changes of the fruit by means of the deep learning method. Deep learning has gained attention in fruit classification and recognition in unimodal processing. This paper proposes a multimodal data fusion framework for the determination of pineapple quality using deep learning methods based on the feature extraction acquired from thermal imaging. Feature extraction was selected from the thermal images that provided a correlation with the quality attributes of the fruit in developing the deep learning models. Three different types of deep learning architectures, including ResNet, VGG16, and InceptionV3, were built to develop the multimodal data fusion framework for the classification of pineapple varieties based on the concatenation of multiple features extracted by the robust networks. The multimodal data fusion coupled with powerful convolutional neural network architectures can remarkably distinguish different pineapple varieties. The proposed multimodal data fusion framework provides a reliable determination of fruit quality that can improve the recognition accuracy and the model performance up to 0.9687. The effectiveness of multimodal deep learning data fusion and thermal imaging has huge potential in monitoring the real-time determination of physicochemical changes of fruit.

Visual Detection of Water Content Range of Seabuckthorn Fruit Based on Transfer Deep Learning.

To realize the classification of sea buckthorn fruits with different water content ranges, a convolution neural network (CNN) detection model of sea buckthorn fruit water content ranges was constructed. In total, 900 images of seabuckthorn fruits with different water contents were collected from 720 seabuckthorn fruits. Eight classic network models based on deep learning were used as feature extraction for transfer learning. A total of 180 images were randomly selected from the images of various water content ranges for testing. Finally, the identification accuracy of the network model for the water content range of seabuckthorn fruit was 98.69%, and the accuracy on the test set was 99.4%. The program in this study can quickly identify the moisture content range of seabuckthorn fruit by collecting images of the appearance and morphology changes during the drying process of seabuckthorn fruit. The model has a good detection effect for seabuckthorn fruits with different moisture content ranges with slight changes in characteristics. The migration deep learning can also be used to detect the moisture content range of other agricultural products, providing technical support for the rapid nondestructive testing of moisture contents of agricultural products.

Application Of The Backpropagation Method For Digital Image Feature Extraction On Coffee Fruit Classification

The background of this research is that the process of classifying the types of coffee beans is still carried out conventionally and has not been computerized, that is, the classification of coffee beans is still based on experience, color and shape of the coffee beans. Of course, this takes a long time and errors often occur, so this research can help classify coffee cherries using the Backpropagation method quickly. The purpose of this study was to produce a system for applying the backpropagation method for extracting digital image features in the classification of coffee cherries, a system built using MATLAB software. In the results of the coffee fruit classification program with the Backpropagation algorithm that is matched in excel, results are obtained for an accuracy of 72% for 100 training data and results for an accuracy of 85% for 10 test data for manual calculations, results are obtained from the Performance Model Using the Confusion Matrix in training data obtained an accuracy of 72% and testing 85% of each of the 2 different types of coffee cherries.

Fruit Image Classification Model Based on MobileNetV2 with Deep Transfer Learning Technique

Due to the rapid emergence and evolution of AI applications, the utilization of smart imaging devices has increased significantly. Researchers have started using deep learning models, such as CNN, for image classification. Unlike the traditional models, which require a lot of features to perform well, CNN does not require any handcrafted features to perform well. It uses numerous filters, which extract required features from images automatically for classification. One of the issues in the horticulture industry is fruit classification, which requires an expert with a lot of experience. To overcome this issue an automated system is required which can classify different types of fruits without the need for any human effort. In this study, a dataset of a total of 26,149 images of 40 different types of fruits was used for experimentation. The training and test set were randomly recreated and divided into the ratio of 3:1. The experiment introduces a customized head of five different layers into MobileNetV2 architecture. The classification layer of the MobileNetV2 model is replaced by the customized head, which produced the modified version of MobileNetV2 called TL-MobileNetV2. In addition, transfer learning is used to retain the pre-trained model. TL-MobileNetV2 achieves an accuracy of 99%, which is 3% higher than MobileNetV2, and the equal error rate of TL-MobileNetV2 is just 1%. Compared to AlexNet, VGG16, InceptionV3, and ResNet, the accuracy is better by 8, 11, 6, and 10%, respectively. Furthermore, the TL-MobileNetV2 model obtained 99% precision, 99% for recall, and a 99% F1-score. It can be concluded that transfer learning plays a big part in achieving better results, and the dropout technique helps to reduce the overfitting in transfer learning.

Recent advances in automatic feature detection and classification of fruits including with a special emphasis on Watermelon (Citrillus lanatus): A review

This document provides an overview of advances in the task of automatic feature detection and classification of fruits, with and special interest in watermelon (Citrillus lanatus). The review was written with the objective to highlight the wealth of knowledge that exist in the application of analytical, smart and sensing image recognition techniques commonly used in agro-industry, and the computational approaches used to make the classification possible. Also, an specific interest is put in the contributions made in the development of automatic recognition systems geared towards for watermelon using images, acoustic and spectroscopy methodologies. The importance of this document is that it provides a conceptual summary of the methods for the automatic recognition of fruits, including machine learning and evolutionary computational algorithms to analyze their sensed data. In conclusion this is a first step into recognizing the challenges and opportunities that can be addressed in this field to augment the visibility of these methods and to further modernize the agro-industrial sector.

The Novel Combination of Nano Vector Network Analyzer and Machine Learning for Fruit Identification and Ripeness Grading

Fruit classification is required in many smart-farming and industrial applications. In the supermarket, a fruit classification system may be used to help cashiers and customer to identify the fruit species, origin, ripeness, and prices. Some methods, such as image processing and NIRS (near-infrared spectroscopy) are already used to classify fruit. In this paper, we propose a fast and cost-effective method based on a low-cost Vector Network Analyzer (VNA) device augmented by K-nearest neighbor (KNN) and Neural Network model. S-parameters features are selected, which take into account the information on signal amplitude or phase in the frequency domain, including reflection coefficient S11 and transmission coefficient S21. This approach was experimentally tested for two separate datasets of five types of fruits, including Apple, Avocado, Dragon Fruit, Guava, and Mango, for fruit recognition as well as their level of ripeness. The classification accuracy of the Neural Network model was higher than KNN with 98.75% and 99.75% on the first dataset, whereas the KNN was seen to be more effective in classifying ripeness with 98.4% as compared to 96.6% for neural network.

Towards sweetness classification of orange cultivars using short-wave NIR spectroscopy

The global orange industry constantly faces new technical challenges to meet consumer demands for quality fruits. Instead of traditional subjective fruit quality assessment methods, the interest in the horticulture industry has increased in objective, quantitative, and non-destructive assessment methods. Oranges have a thick peel which makes their non-destructive quality assessment challenging. This paper evaluates the potential of short-wave NIR spectroscopy and direct sweetness classification approach for Pakistani cultivars of orange, i.e., Red-Blood, Mosambi, and Succari. The correlation between quality indices, i.e., Brix, titratable acidity (TA), Brix: TA and BrimA (Brix minus acids), sensory assessment of the fruit, and short-wave NIR spectra, is analysed. Mix cultivar oranges are classified as sweet, mixed, and acidic based on short-wave NIR spectra. Short-wave NIR spectral data were obtained using the industry standard F-750 fruit quality meter (310–1100 nm). Reference Brix and TA measurements were taken using standard destructive testing methods. Reference taste labels i.e., sweet, mix, and acidic, were acquired through sensory evaluation of samples. For indirect fruit classification, partial least squares regression models were developed for Brix, TA, Brix: TA, and BrimA estimation with a correlation coefficient of 0.57, 0.73, 0.66, and 0.55, respectively, on independent test data. The ensemble classifier achieved 81.03% accuracy for three classes (sweet, mixed, and acidic) classification on independent test data for direct fruit classification. A good correlation between NIR spectra and sensory assessment is observed as compared to quality indices. A direct classification approach is more suitable for a machine-learning-based orange sweetness classification using NIR spectroscopy than the estimation of quality indices.

High Zoom Ratio Foveated Snapshot Hyperspectral Imaging for Fruit Pest Monitoring

Snapshot hyperspectral imaging technology is increasingly used in agricultural product monitoring. In this study, we present a 9× local zoom snapshot hyperspectral imaging system. Using commercial spectral sensors with spectrally resolved detector arrays, we achieved snapshot hyperspectral imaging with 14 wavelength bands and a spectral bandwidth of 10–15 nm. An experimental demonstration was performed by acquiring spatial and spectral information about the fruit and Drosophila. The results show that the system can identify Drosophila and distinguish well between different types of fruits. The results of this study have great potential for online fruit classification and pest identification.

Fruit Classification using Colorized Depth Images

Fruit classification is a computer vision task that aims to classify fruit classes correctly, given an image. Nearly all fruit classification studies have used RGB color images as inputs, a few have used costly hyperspectral images, and a few classical ML-based have used colorized depth images. Depth images have apparent benefits such as invariance to lighting, less storage requirement, better foreground-background separation, and more pronounced curvature details and object edge discontinuities. However, the use of depth images in CNN-based fruit classification remains unexplored. The purpose of this study is to investigate the use of colorized depth images in fruit classification with four CNN models, namely, AlexNet, GoogleNet, ResNet101, and VGG16, and compare their performance and computational efficiency, as well as the impact of transfer learning. Depth images of apple, orange, mango, banana and rambutan (Nephelium Lappaceum) were manually collected using a depth sensor with sub-millimeter accuracy and subjected to jet, uniform, and inverse colorization to produce three sets of dataset. Results show that depth images can be used to train CNN models for fruit classification with ResNet101 achieving the best accuracy of 96%on the inverse dataset. It achieved 100% accuracy after transfer learning. GoogleNet showed the most significant improvement after transfer learning on the uniform dataset, at 12.27%. It also exhibited the lowest training and inference times. The results show the potential use of depth images for fruit classification and similar computer vision tasks.

LightNet: pruned sparsed convolution neural network for image classification

Deep learning has become the most sought-after approach in the area of artificial intelligence (AI). However, deep learning models pose some challenges in the learning process. It is computationally intensive to train deep learning networks and also resource-intensive. Therefore, it cannot be applicable in limited-resource devices. Limited research is being done on the implementation of efficient approaches for real-world problems. This study tries to bridge the gaps towards an applicable system in the real world especially in the agricultural sector for plants disease management and fruits classification. We introduce a novel deep learning architecture called LightNet. LightNet is an architecture that employs two strategies to achieve the sparsity of DenseNet: the skip connections and pruning strategy. The resultant is a small network with reduced parameters and model size. LightNet model matches the performance of the highest accuracy of DenseNet. Moreover, LightNet is × 2 smaller, × 2 parameters efficient and × 3 faster compared original DenseNet. The model is evaluated on the real-worlds dataset PlantsVillage and Fruits-360. The results show that our model achieved state-of-the-art results and it can be used for plant disease detection and fruits classification and grading.

Segregation of Ripe and Raw Bananas Using Convolutional Neural Network

Agricultural sector is one of the main sectors in India. Productive growth and high yield production of fruits is important and required for the agricultural industry. Application of computer vision and image processing has aid agriculture to boost yield rating, disease spotting, irrigation, fruit classification and maturity grading. Computer vision and image processing approach can be used to lower the time utilization and has made it economical. A convolutional neural network (CNN) is a type of artificial neural network used in image identification and processing that is mainly planned to process pixel data. This paper utilized a convolutional neural network (CNN) to classify bananas into raw or ripe without requiring labour. Bananas were classified by their colour appearance in the CNN classification and not only this it can be also used to classify many other types of fruits and vegetables.

A Hybrid Particle Size Algorithm for Classification of Hygienic Fruit and Vegetable Images Based on Convolution Neural Network from Health Perspective

Abstract In order to improve the clarity of selection of hygienic fruit and vegetable images, a hybrid granularity classification algorithm for fruit and vegetable images based on convolutional neural network is proposed. The edges of fruit and vegetable images are detected, and the fruit and vegetable images are preprocessed under the convolutional neural network. The fruit and vegetable images are sampled in the form of equal intervals, the mixed noise in the fruit and vegetable images is judged, and the window pixels are marked. Finally, the wavelet threshold algorithm is used to filter the noise of the fruit and vegetable image, and the wavelet of the effective low-frequency signal is reconstructed combined with the convolution neural network the mixed noise filtered fruit and vegetable image is obtained by using the threshold high frequency signal coefficient and the threshold high frequency signal coefficient. Experimental results show that the algorithm has high definition, good denoising effect, and high measurement accuracy.

Provisional Classification of Wild Edible Fruits Based on their Nutrient Profiles

Provisional Classification of Wild Edible Fruits Based on their Nutrient Profiles

Artificial Humming Bird Optimization with Siamese Convolutional Neural Network Based Fruit Classification Model

Artificial Humming Bird Optimization with Siamese Convolutional Neural Network Based Fruit Classification Model

Fruit Classification Based on External Defects Using Deep Learning

Fruit Classification Based on External Defects Using Deep Learning

A cross-domain fruit classification method based on lightweight attention networks and unsupervised domain adaptation

AbstractImage-based fruit classification offers many useful applications in industrial production and daily life, such as self-checkout in the supermarket, automatic fruit sorting and dietary guidance. However, fruit classification task will have different data distributions due to different application scenarios. One feasible solution to solve this problem is to use domain adaptation that adapts knowledge from the original training data (source domain) to the new testing data (target domain). In this paper, we propose a novel deep learning-based unsupervised domain adaptation method for cross-domain fruit classification. A hybrid attention module is proposed and added to MobileNet V3 to construct the HAM-MobileNet that can suppress the impact of complex backgrounds and extract more discriminative features. A hybrid loss function combining subdomain alignment and implicit distribution metrics is used to reduce domain discrepancy during model training and improve model classification performance. Two fruit classification datasets covering several domains are established to simulate common industrial and daily life application scenarios. We validate the proposed method on our constructed grape classification dataset and general fruit classification dataset. The experimental results show that the proposed method achieves an average accuracy of 95.0% and 93.2% on the two datasets, respectively. The classification model after domain adaptation can well overcome the domain discrepancy brought by different fruit classification scenarios. Meanwhile, the proposed datasets and method can serve as a benchmark for future cross-domain fruit classification research.

Innovative Models Built Based on Image Textures Using Traditional Machine Learning Algorithms for Distinguishing Different Varieties of Moroccan Date Palm Fruit (Phoenix dactylifera L.)

The aim of this study was to develop the procedure for the varietal discrimination of date palm fruit using image analysis and traditional machine learning techniques. The fruit images of ‘Mejhoul’, ‘Boufeggous’, ‘Aziza’, ‘Assiane’, and ‘Bousthammi’ date varieties, converted to individual color channels, were processed to extract the texture parameters. After performing the attribute selection, the textures were used to build models intended for the discrimination of different varieties of date palm fruit using machine learning algorithms from Functions, Bayes, Lazy, Meta, and Trees groups. Models were developed for combining image textures selected from a set of all color channels and for sets of textures selected for individual color spaces and color channels. The models, including combined textures selected from all color channels, distinguished all five varieties with an average accuracy reaching 98%, and ‘Bousthammi’ and ‘Mejhoul’ were completely correctly discriminated for the SMO (Functions) and IBk (Lazy) machine learning algorithms. By reducing the number of varieties, the correctness of the date palm fruit classification increased. The models developed for the three most different date palm fruit varieties ‘Boufeggous’, ‘Bousthammi’, and ‘Mejhoul’ revealed an average discrimination accuracy of 100% for each algorithm used (SMO, Naive Bayes (Bayes), IBk, LogitBoost (Meta), and LMT (Trees)). In the case of individual color spaces and channels, the accuracies were lower, reaching 97.3% for color space RGB and SMO and LMT algorithms for all five varieties and 99.63% for Naive Bayes and IBk for the ‘Boufeggous’, ‘Bousthammi’, and ‘Mejhoul’ date palm fruits. The results can be used in practice to develop vision systems for sorting and distinguishing the varieties of date palm fruit to authenticate the variety of the fruit intended for further processing.

Automatic Identification of banana quality with Deep Neural Network Classification (DNN)

The industrialization of agriculture has gradually taken place in Colombia, because of the need to optimize the time and effort required for the processes involved from growing fruits and vegetables to their distribution and commercialization, whether they are meant for exportation or commercialization within the country. Fruit classification is a very important process before they are being taken out as a final product, since predetermined guidelines issued by regulatory entities that define the quality of the harvested product must be followed. The authors propose a low-cost prototype for the automatic classification of bananas according to the NTC 1190 standard (Colombian normative), using a convolutional neural network (CNN) of MobileNetV2 architecture trained through transfer learning and implemented in a Raspberry Pi 3B+ with a camera to monitor the specimens and an easy interface for interaction with the user, as well as a case designed to contain the hardware and allow access to its ports in the most compact way possible. The datasets utilized in this work for training, validation, and testing, consists of images taken from two free access fruit database and others acquired by the researchers. The achieved precision is 87 %, enough to ensure reliability and low computational cost.

Laboratory and Field Trials to Identify Reduced-Risk Insecticides for the Control of the Golden Twin-Spot Moth Chrysodeixis chalcites (Esper) (Lepidoptera: Noctuidae) in Banana Plantations

The golden twin-spot moth (Chrysodeixis chalcites) is one of the most important pests in banana production on the Canary Islands (Spain). The efficacy of different biorational insecticides based on bioenzyme complexes (Intruder®), plant extracts: Rutaceae and Piperaceae (Avenger®), Rutaceae and Lauraceae (BioKnock®), cinnamon, citronella, and Menta (Cinamite®), Alliaceae and Solanacea (Garlitrol-Forte®), citrus (Prevam®), and neem oil (Indasol®) was assessed against C. chalcites. Laboratory assays included: choice (repellent effect), no choice, and contact toxicity on C. chalcites 2nd instar larvae. The highest repellent effect was observed with Prevam® (85.19 ± 1.7%), followed by Garlitrol® (68.44 ± 5.7%) and Intruder® (67.54 ± 4.3%). In no choice assays, Prevam® (0.92 ± 0.4%), Indasol® (0.98 ± 0.33%), and Intruder® (2.7 ± 0.33%) had the lowest leaf consumption. The contact toxicity assays showed the highest mortality with Intruder® both at 1 day and 7 days post-application (20.22 ± 2.98% and 77.77 ± 5.7%, respectively). In the screenhouse trial, the best results for C. chalcites larvae mortality, fruit damage, and fruit classification in quality categories 7 days after application of the bioinsecticide were obtained with Intruder®, Prevam®, and Indasol®. An economic analysis of biorational treatments was also performed. The results of this study provide successful alternatives to chemical pesticides for the control of C. chalcites on banana plants in the Canary Islands.

Reliable classification of nutmeg fruits from some geographical origins using SWNIR spectroscopy associated with PLS-DA and SVM methods: Case Study in Aceh Province, Indonesia

Essential oils of nutmeg (Myristica fragrans) can have different qualities due to the soil and industrial process, as Indonesia’s main nutmeg production areas are enormous, especially in Aceh province. The lack of information on detecting chemical constituents and classifying nutmeg from Aceh province makes the price of nutmeg from Aceh unable to compete. Therefore, this study aims to reliably classify nutmeg fruits of some geographical origins in Aceh Province using shortwave near-infrared (SWNIR) spectroscopy associated with partial least squares discriminant analysis (PLS-DA) and support vector machine (SVM). The SWNIR used in this study has a wavelength of 381 to 1065 nm at a resolution of 3 nm and a diffuse reflectance mode. The fruit comes from four geographical areas of origin in the province of Aceh. A total of forty SWNIR spectral data were further classified using two algorithms, namely PLS-DA and SVM. The results show that SWNIR can correctly classify nutmeg from some sub-districts in Aceh province with the help of the PLS-DA and SVM algorithms. From this study, the following SVM algorithm is recommended to be used to classify nutmeg from the province of Aceh with an outstanding level of precision.