The database used in this study is comprising of 44406 fruit images, which we collected in a period of 6 months. The images where made with in our lab’s environment under different scenarios which we mention below. We captured all the images on a clear background with resolution of 320×258 pixels. We used HD Logitech web camera to took the pictures. During collecting this database, we created all kind of challenges, which, we have to face in real-world recognition scenarios in supermarket and fruit shops such as light, shadow, sunshine, pose variation, to make our model robust for, it might be necessary to cope with illumination variation, camera capturing artifacts, specular reflection shading and shadows. We tested our model’s robustness in all scenarios and it perform quit well.

Fruit Recognition

## Overview

Fruit Recognition is a dataset for object detection tasks - it contains Fruits annotations for 1,649 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

This dataset, initially consisting of 10,154 high-resolution images of five fruit types—apple, banana, mango, orange, and grapes—has been expanded to over 81,000 using advanced augmentation techniques like rotation, scaling, and brightness adjustment. The images are classified into three key categories: fresh, rotten, and formalin-mixed. This dataset offers a unique opportunity for researchers in computer vision, agriculture, and food safety to develop machine learning and deep learning models for tasks such as real-time fruit quality assessment, contamination detection, and automating food inspection processes. Its extensive and diverse collection makes it a valuable resource for innovations in public health, export quality control, and agricultural productivity.

(1) Everyone seeks fresh and high-quality fruits, as fruits naturally degrade over time, leading to spoilage. It is estimated that roughly one-third of harvested fruits rot, causing significant financial loss. Additionally, the sale of fruits is impacted by consumer concerns over the health risks associated with consuming spoiled or chemically treated fruits. The manual classification of fresh and rotten fruits by individuals is inefficient, particularly for farmers, sellers, and the fruit processing industry. (2) In recent years, computer vision techniques have shown great promise in automating tasks like classification and detection of fresh and rotten fruits. (3) To aid in the development of computer vision-based algorithms, we present an extensive fruit dataset containing five fruit classes: Fresh Apple, Rotten Apple, Formalin-mixed Apple, Fresh Banana, Rotten Banana, Formalin-mixed Banana, Fresh Grape, Rotten Grape, Formalin-mixed Grape, Fresh Mango, Rotten Mango, Formalin-mixed Mango, Fresh Orange, Rotten Orange, and Formalin-mixed Orange. Classification into fresh, rotten, and formalin-treated categories was carried out with the assistance of agricultural experts.

Here are a few use cases for this project:

1. Farm Automation: The model could be used in smart farming applications to identify and categorize fruits during harvesting. This could help automate the harvesting process and increase the overall productivity of the farm.

2. Grocery Store Management: The model could be integrated into inventory management systems at grocery stores to automatically identify and count various types of fruits. This would simplify stock tracking, aid in automatic pricing, and reduce manual labor costs.

3. Health and Nutrition Apps: The model could be incorporated into health and nutrition apps aimed at helping users track their food intake. Users could take pictures of their meals, and the app could identify any fruit present, aiding in the calculation of nutritional intake.

4. Food Safety and Quality Inspection: The model could be used in food processing industries to identify and remove non-fruit objects or unfit fruits from the production line, thus ensuring the quality and safety of the products.

5. Educational Tools: The model could be included in interactive educational software or mobile apps aimed at teaching children about different types of fruits and their nutritional benefits.

This dataset is a customized version of the Fruit Quality Classification dataset originally created by Ryan D. Park (https://www.kaggle.com/datasets/ryandpark/fruit-quality-classification), which provides detailed images of different fruits categorized by quality and type. I have adapted and expanded upon this dataset to create a unique collection focused specifically on detection across various fruit types.

Dataset Overview: - Total Images: 1,968 images - Training Set: 1,852 images - Validation Set: 116 images

Fruit Types:Apple, Banana, Guava, Lime, Orange, Pomegranate Quality Categories: This dataset categorizes fruits into different ripeness stages, capturing a wide range of ripening conditions: - Bad Quality - Good Quality

Use Case: This dataset can be particularly useful for: - Fruit Quality Classification: Aiming to distinguish between quality in fruits, which has applications in agricultural technology, quality control, and the food industry. - AI and Computer Vision Projects: Perfect for training deep learning models to detect ripeness levels from fruit images. - Agricultural Research: Facilitates research into automating the identification of fruit quality and ripeness, a key factor in food logistics and waste reduction.

Dataset Details: Image Dimensions: Images are primarily of size 256x256 pixels, sourced from diverse real-world environments, captured under varying lighting conditions and backgrounds (e.g., top views, front views, rotated orientations). Annotations: Each image is labeled with bounding box coordinates for detecting fruits, facilitating object detection tasks as well as classification of ripeness levels.

Credits: This dataset was developed using images from the publicly available Fruit Quality Classification dataset by Ryan D. Park (https://www.kaggle.com/datasets/ryandpark/fruit-quality-classification). All original credits go to the creator, and I extend my thanks for sharing this dataset. The original dataset contains a much larger collection of images across different fruit types and quality levels, which served as the basis for this specialized ripeness-focused version.

About Dataset (strawberries, peaches, pomegranates) Photo requirements: 1-White background 2-.jpg 3- Image size 300*300 The number of photos required is 250 photos of each fruit when it is fresh and 250 photos of each Fruit Dataset for Classification when it is rotten. Total 1500 images

Diverse Collection With a diverse collection of Product images, the files provides an excellent foundation for developing and testing machine learning models designed for image recognition and allocation. Each image is captured under different lighting conditions and backgrounds, offering a realistic challenge for algorithms to overcome.

Real-World Applications The variability in the dataset ensures that models trained on it can generalize well to real-world scenarios, making them robust and reliable. The dataset includes common fruits such as apples, bananas, oranges, and strawberries, among others, allowing for comprehensive training and evaluation.

Industry Use Cases One of the significant advantages of using the Fruits Dataset for Classification is its applicability in various fields such as agriculture, retail, and the food industry. In agriculture, it can help automate the process of fruit sorting and grading, enhancing efficiency and reducing labor costs. In retail, it can be used to develop automated checkout systems that accurately identify fruits, streamlining the purchasing process.

Educational Value The dataset is also valuable for educational purposes, providing students and educators with a practical tool to learn and teach machine learning concepts. By working with this dataset, learners can gain hands-on experience in data preprocessing, model training, and evaluation.

Conclusion The Fruits Dataset for Classification is a versatile and indispensable resource for advancing the field of image classification. Its diverse and high-quality images, coupled with practical applications, make it a go-to dataset for researchers, developers, and educators aiming to improve and innovate in machine learning and computer vision.

This dataset is sourced from Kaggle.

Fruits By YOLO

## Overview

Fruits By YOLO is a dataset for object detection tasks - it contains Fruits annotations for 1,176 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

This dataset was curated for developing CNN models for binary classification in fruit recognition. It comprises images belonging to two classes: Apples and Mangos. The dataset is sourced from the Fruit-360 dataset.

High quality images of fruits are required to solve fruit classification and recognition problem. To build the machine learning models, neat and clean dataset is the elementary requirement. With this objective we have created the dataset of six popular Indian fruits named as “FruitNet”. This dataset consists of 14700+ high-quality images of 6 different classes of fruits in the processed format. The images are divided into 3 sub-folders 1) Good quality fruits 2) Bad quality fruits and 3) Mixed quality fruits. Each sub-folder contains the 6 fruits images i.e. apple, banana, guava, lime, orange, and pomegranate. Mobile phone with a high-end resolution camera was used to capture the images. The images were taken at the different backgrounds and in different lighting conditions. The proposed dataset can be used for training, testing and validation of fruit classification or reorganization model.

[The related article is available at: https://www.sciencedirect.com/science/article/pii/S2352340921009616. Cite the article as : V. Meshram, K. Patil, FruitNet: Indian fruits image dataset with quality for machine learning applications, Data in Brief, Volume 40, 2022, 107686, ISSN 2352-3409, https://doi.org/10.1016/j.dib.2021.107686 ]

Fruit Object Detection is a dataset for an object detection task. Possible applications of the dataset could be in the food industry. The dataset consists of 4474 images with 22576 labeled objects belonging to 11 different classes including pear, apple, grape, and other: pineapple, durian, korean melon, watermelon, tangerine, lemon, cantaloupe, and dragon fruit

Credit: 'Fruit recognition from images using deep learning' by H. Muresan and M. Oltean (https://arxiv.org/abs/1712.00580)

Fruit 360 is a dataset with 90380 images of 131 fruits and vegetables (https://www.kaggle.com/moltean/fruits). Images are 100 pixel by 100 pixel and are RGB (color) images (3 values for each pixel). This dataset is a subset of Fruit 360 dataset, containing only 10 fruits/vegetables (Strawberry, Apple_Red_Delicious, Pepper_Green, Corn, Banana, Tomato_1, Potato_White, Pineapple, Orange, and Peach). We selected a subset of fruits/vegetables, so the dataset size is smaller and the neural network can be trained faster.

The utilities used to create the dataset, along with step by step instructions, can be found here: https://github.com/kxk302/fruit_dataset_utilities

First, we created feature vectors for each image. Each image is 100 pixel by pixel and are RGB (color) images (3 values for each pixel). Hence, each image can be represented by 30,000 values (100 X 100 X 3). Second, we selected a subset of 10 fruits/vegetables images (training and test dataset sizes go from 7 GB and 2.5 GB for 131 fruits/vegetables to 500 MB and 177 MB for 10 fruits/vegetables, respectively). Third, we created separate files for feature vectors and labels. Finally, we mapped the labels for the 10 selected fruits/vegetables to a range of 0 to 9.

Fruit Detection

## Overview

Fruit Detection is a dataset for object detection tasks - it contains Fruits annotations for 1,155 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [BY-NC-SA 4.0 license](https://creativecommons.org/licenses/BY-NC-SA 4.0).

The "Date Fruit Detection Dataset for Computer Vision-Based Automatic Harvesting" is a collection of videos and images showcasing date fruits from four Moroccan varieties, namely Majhoul, Boufagous, Bouisthami, and Khoult. This dataset is specifically designed to detect and classify date fruits, with the primary goal being the automation of the harvesting process. All the images in this dataset were captured in two orchards located in Morocco, with the first orchard situated in the southeast of Errachidia and the second in Tismoumine, Alnif, Tinghir. These images were taken under various natural conditions, encompassing differing lighting, contrast, shadows, and instances where the dates were concealed by bags or hidden beneath palm leaves. The dataset was meticulously compiled over the period spanning from June to September 2022, ensuring comprehensive coverage of all four maturity stages of date fruits, which include immature, khalal, rutab, and tamer. The dataset is intended for both object detection and classification purposes, and it includes a YOLO annotation txt file for each image. These annotations have been tailored to precisely recognize not only the date fruit but also to distinguish the specific variety and its maturity stage.

The Nature3 dataset offers annotated images of plant leaves, flowers, and fruits for real-time detection using YOLO models, ideal for agriculture, plant health monitoring, and environmental research.

The dataset comprises images of fruits from 12 different categories. Each category contains a diverse set of images to ensure the robustness and generalizability of the trained Convolutional Neural Network (CNN) model. The images have been collected and curated to represent various conditions such as different lighting, angles, and backgrounds.

Categories The dataset includes the following 12 fruit categories:

Apple Banana Cherry Grape Kiwi Mango Orange Peach Pear Plum Strawberry Watermelon

Collection: Images were sourced from various online repositories and manually verified to ensure quality and relevance.

Usage To use the dataset for training the CNN model, simply organize the images as described above and load them using your preferred deep learning library (e.g., Fast.ai, TensorFlow, PyTorch).

About the Dataset

Context

This dataset encompasses images of various fruits and vegetables, providing a diverse collection for image recognition tasks. The included food items are:

• Fruits: Banana, Apple, Pear, Grapes, Orange, Kiwi, Watermelon, Pomegranate, Pineapple, Mango
• Vegetables: Cucumber, Carrot, Capsicum, Onion, Potato, Lemon, Tomato, Radish, Beetroot, Cabbage, Lettuce, Spinach, Soybean, Cauliflower, Bell Pepper, Chilli Pepper, Turnip, Corn, Sweetcorn, Sweet Potato, Paprika, Jalapeño, Ginger, Garlic, Peas, Eggplant

Content

The dataset is organized into three main folders: - Train: Contains 100 images per category. - Test: Contains 10 images per category. - Validation: Contains 10 images per category.

Each of these folders is subdivided into specific folders for each type of fruit and vegetable, containing respective images.

Data Collection

The images in this dataset were sourced using Bing Image Search for a personal project focused on image recognition of food items. The creator does not hold the rights to any of the images included in this dataset. If you are the owner of any image and have concerns regarding its use, please contact the creator to request its removal. The creator will promptly comply with any such requests to ensure all legal obligations are met.

Disclaimer: Users of this dataset are responsible for ensuring that their use of the images complies with applicable copyright laws and regulations. The creator assumes no responsibility for any legal issues that may arise from the use of this dataset. It is recommended to use the dataset for educational and non-commercial purposes only and to seek legal counsel if you have specific concerns about copyright compliance.

Inspiration

The primary motivation behind creating this dataset was to develop an application capable of recognizing food items from photographs. The application aims to suggest various recipes that can be prepared using the identified ingredients.

Citation

Kritik Seth, "Fruits and Vegetables Image Recognition Dataset," Kaggle 2020 [https://www.kaggle.com/kritikseth/fruit-and-vegetable-image-recognition]

Vege And Fruit Detection

## Overview

Vege And Fruit Detection is a dataset for object detection tasks - it contains Fruits annotations for 1,062 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

This is the dataset associated with MDPI Sensors paper entitled "DeepFruits: A Fruit Detection System Using Deep Neural Networks".

This paper presents a novel approach to fruit detection using deep convolutional neural networks. The aim is to build an accurate, fast and reliable fruit detection system, which is a vital element of an autonomous agricultural robotic platform; it is a key element for fruit yield estimation and automated harvesting. Recent work in deep neural networks has led to the development of a state-of-the-art object detector termed Faster Region-based CNN (Faster R-CNN). We adapt this model, through transfer learning, for the task of fruit detection using imagery obtained from two modalities: colour (RGB) and Near-Infrared (NIR). Early and late fusion methods are explored for combining the multi-modal (RGB and NIR) information. This leads to a novel multi-modal Faster R-CNN model, which achieves state-of-the-art results compared to prior work with the F1 score, which takes into account both precision and recall performances improving from 0.807 to 0.838 for the detection of sweet pepper. In addition to improved accuracy, this approach is also much quicker to deploy for new fruits, as it requires bounding box annotation rather than pixel-level annotation (annotating bounding boxes is approximately an order of magnitude quicker to perform). The model is retrained to perform the detection of seven fruits, with the entire process taking four hours to annotate and train the new model per fruit.

This dataset consists of four sets of flower images, from three different species: apple, peach, and pear, and accompanying ground truth images. The images were acquired under a range of imaging conditions. These datasets support work in an accompanying paper that demonstrates a flower identification algorithm that is robust to uncontrolled environments and applicable to different flower species. While this data is primarily provided to support that paper, other researchers interested in flower detection may also use the dataset to develop new algorithms. Flower detection is a problem of interest in orchard crops because it is related to management of fruit load. Funding provided through ARS Integrated Orchard Management and Automation for Deciduous Tree Fruit Crops. Resources in this dataset:Resource Title: AppleA images. File Name: AppleA.zipResource Description: 147 images of an apple tree in bloom acquired with a Canon EOS 60D.Resource Title: Training image names from Apple A dataset. File Name: train.txtResource Description: This is a list of filenames used in training; see related paper for details.Resource Title: AppleA labels. File Name: AppleA_Labels.zipResource Description: Binary images for the Apple A set, where white represents flower pixels and black, non-flower pixels. June 25, 2018: 5 files added: 275.png, 316.png, 328.png, 336.png, 369.png.Resource Title: Validation image names from Apple A dataset. File Name: val.txtResource Description: This is a list of filenames used in testing; see related paper for details. June 25, 2018: 5 filenames added. IMG_0275.JPG IMG_0316.JPG IMG_0328.JPG IMG_0336.JPG IMG_0369.JPGResource Title: AppleB images. File Name: AppleB.zipResource Description: 15 images of an apple tree in bloom acquired with a GoPro HERO 5. June 25, 2018: 3 files added. 23.bmp 28.bmp 42.bmpResource Title: AppleB labels. File Name: AppleB_Labels.zipResource Description: Binary images for the Apple B set, where white represents flower pixels and black, non-flower pixels. June 25, 2018: 3 files added. 23.bmp 28.bmp 42.bmpResource Title: Peach. File Name: Peach.zipResource Description: 20 images of an peach tree in bloom acquired with a GoPro HERO 5. June 25, 2018: 4 files added. 14.bmp 34.bmp 40.bmp 41.bmpResource Title: Peach labels. File Name: PeachLabels.zipResource Description: Binary images for the Peach set, where white represents flower pixels and black, non-flower pixels. June 25, 2018: 4 files added. 14.bmp 34.bmp 40.bmp 41.bmpResource Title: Pear. File Name: Pear.zipResource Description: 15 images of a free-standing pear tree in bloom, acquired with a GoPro HERO5. June 25, 2018: 3 files added. 1_25.bmp 1_62.bmp 2_28.bmpResource Title: Pear labels. File Name: PearLabels.zipResource Description: Binary images for the pear set, where white represents flower pixels and black, non-flower pixels. June 25, 2018: 3 files added. 1_25.bmp 1_62.bmp 2_28.bmpResource Title: Apple A Labeled images from training set . File Name: AppleALabels_Train.zipResource Description: Binary images for the Apple A set, where white represents flower pixels and black, non-flower pixels. These images form the training set. Resource added August 20, 2018. User noted that this resource was missing.

Fresh And Rotten Fruit Detection

## Overview

Fresh And Rotten Fruit Detection is a dataset for object detection tasks - it contains Fruits annotations for 561 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).