Abstract:The fruit identification process involves analyzing and categorizing different types of fruits based on their visual characteristics. This activity can be achieved using a range of methodologies, encompassing manual examination, conventional computer vision methodologies, and more sophisticated methodologies employing machine learning and deep learning. Our study identified a total of 15 distinct categories of fruit, consisting of class Avocado, Banana, Cherry, Apple Braeburn, Apple golden 1, Apricot, Grape, Kiwi, Mango, Orange, Papaya, Peach, Pineapple, Pomegranate and Strawberry. Neural Architecture Search (NAS) is a technological advancement employed within the realm of deep learning and artificial intelligence, to automate conceptualizing and refining neural network topologies. NAS aims to identify neural network structures that are highly suitable for tasks, such as the detection of fruits. Our suggested model with 99.98% mAP increased the detection performance of the preceding research study that used Fruit datasets. In addition, after the completion of the study, a comparative analysis was carried out to assess the findings in conjunction with those of another research that is connected to the topic. When compared to the findings of earlier studies, the detector that was proposed exhibited higher performance in terms of both its accuracy and its precision.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the classification and detection problems in the fruit recognition process. Specifically, researchers are concerned with how to use deep learning and Neural Architecture Search (NAS) techniques to improve the accuracy and efficiency of fruit recognition. It is mentioned in the paper that traditional fruit recognition methods include manual inspection, traditional computer vision methods, and more advanced machine - learning methods, but these methods have certain limitations when dealing with large - scale and diverse fruit image data sets. Therefore, this paper proposes a new fruit classification system based on NAS, aiming to optimize the fruit detection task by automatically designing the neural network structure, thereby improving the detection performance. The main contributions of the research are as follows: 1. **Introduced an innovative fruit classification system**: Utilized computer vision technology to solve the limitations of existing methods. 2. **Applied NAS technology**: Applied NAS to the fruit classification problem for the first time, improving the adaptability and generalization ability of the model. 3. **Significantly improved detection performance**: Compared with the fruit data sets used in previous studies, the proposed model has a 7% improvement in mAP (mean Average Precision), increasing from 93.78% to 99.98%. 4. **Wide recognition of fruit varieties**: The research covers 15 different fruit categories and can effectively recognize a variety of fruits. 5. **Combined traditional features and texture features**: Not only color and shape features were extracted, but also important texture features were added, further improving the accuracy of the classifier. Through these improvements, this research provides a more efficient and accurate solution for fruit detection in multiple fields such as agriculture, food processing, and retail.

Fruit Classification System with Deep Learning and Neural Architecture Search

DeepFruits: A Fruit Detection System Using Deep Neural Networks

A general purpose multi-fruit system for assessing the quality of fruits with the application of recurrent neural network

Apple Ripeness Identification Using Deep Learning

Deep neural network based fruit identification and grading system for precision agriculture

Fruit recognition from images using deep learning applications

Fruits Classification and Detection Application Using Deep Learning

Recent Advancements in Fruit Detection and Classification Using Deep Learning Techniques

Fruit classification using attention-based MobileNetV2 for industrial applications

Deep Learning-Based Fruit Recognition and Quality Assessment: A Convolutional Neural Network Approach

Fruit Detection from Digital Images Using CenterNet

Fruit Detection and Three-Stage Maturity Grading Using CNN

Transformer based fruits disease classification

YOLO Models for Fresh Fruit Classification from Digital Videos

Implementation of Fruits Recognition Classifier using Convolutional Neural Network Algorithm for Observation of Accuracies for Various Hidden Layers

Research on Apple Recognition Algorithm in Complex Orchard Environment Based on Deep Learning

Fruit Recognition System Using MATLAB

AN APPLE GRADING SYSTEM ACCORDING TO EUROPEAN FRUIT QUALITY STANDARDS USING GABOR FILTER AND ARTIFICIAL NEURAL NETWORKS

Fruit Quality Assessment with Densely Connected Convolutional Neural Network

Fruit Freshness Evaluation using a Real-Time Industrial Framework for Deep Learning Ensemble Approaches

Computer Vision System for Mango Fruit Defect Detection Using Deep Convolutional Neural Network