Mendeley Data Showcase

The Rose Image Classification dataset contains 445 high-quality images of roses, categorized by color. This dataset is ideal for machine learning tasks involving image classification, recognition, and other computer vision applications. Each image is labeled according to the color of the rose, allowing for the development of models that can accurately classify rose varieties based on visual features. Number of Images: 445 Color: Red Roses: 201 images Yellow Roses: 58 images White Roses: 85 images Pink Roses: 93 images

This dataset contains 37,957 high-resolution images of mango fruits collected in Eastern Uganda (Soroti District: Aloet and Madera areas) using smartphone cameras under natural daylight conditions. It includes both healthy and defective mangoes, representing a wide range of post-harvest conditions encountered during harvesting, handling, and marketing. Dataset Organization: Original: 4,699 raw images captured in the field. Preprocessed: 5,064 images resized and center-cropped for consistent framing. Augmented: 28,194 images generated via flipping, brightness adjustment, cropping, and controlled 90°/270° rotations to simulate natural orientation changes. File Format & Access: Images are stored in JPEG format. The dataset is provided in .zpaq format for maximum compression and can be extracted using PeaZip (Windows). A .zip version of the dataset is also available on Kaggle: https://www.kaggle.com/datasets/joanitanamuyiga/mango-fruit-image-classification-uganda Potential Applications: Computer vision and image processing research Fruit quality assessment and post-harvest defect studies Machine learning applications for classification, defect detection, and infection segmentation

This is an image dataset used in our paper The Hidden Cost of the Edge: A Performance Comparison of Edge and Cloud Latencies (to appear in SC21). In our experiments it's used as the workload of a deep neural network (DNN) image classification application.

This CNN model has been written and tested on Google Collab. It can be used for classification of compositional patterns containing color. It has been trained and tested on datasets containing paintings, graphic patterns, photography, interior architecture and building facades providing reliable outcomes. The model employs a ResNet network and can be run in any system as it needs the least space for being loaded and run. It also functions quite fast. The author suggests to use more than 10 epochs when training for better results. The model also can be merged with a similar structured MobileNet model.

Deep learning applied to raw data has demonstrated outstanding image classification performance, mainly when abundant data is available. However, performance significantly degrades when a substantial volume of data is unavailable. Furthermore, in situations where distinguishing between distinct classes is challenging, such as in fine-grained image classification, deep architectures struggle to achieve satisfactory performance levels. Utilizing a priori knowledge alongside raw data can enhance image classification in demanding scenarios. Nevertheless, only a limited number of image classification datasets given with a priori knowledge are currently available, thereby restricting research efforts in this field. This paper introduces innovative datasets for the classification problem that integrate a priori knowledge. These datasets are built from existing data typically employed for multilabel multiclass classification or object detection. Frequent closed itemset mining is used to build classes and their corresponding attributes (e.g. presence of an object in an image) and then to extract a priori knowledge expressed by rules on these attributes. The algorithm for generating rules is described.

Image tiles were created by processing the Landsat 8 Satellite images for the coastal region of Maharashtra to classify Mangroves and Non-Mangroves.

Multi-class weather dataset(MWD) for image classification is a valuable dataset used in the research paper entitled “Multi-class weather recognition from still image using heterogeneous ensemble method”. The dataset provides a platform for outdoor weather analysis by extracting various features for recognizing different weather conditions.

This dataset contains waste images collected from the Primakara University campus environment in Denpasar, Bali, Indonesia. The dataset was developed to support a research project titled “Implementation of an Automatic Waste Sorting System for Organic and Inorganic Waste Based on Artificial Intelligence. The images were captured manually using a smartphone camera with a top-down angle

1200 images were collected in the field in July 2023, in a 5-day window around flowering time of different genotypes. There were three N fertilization levels: N0, N69 and NFull, with no added N, 69 kg of added N and full fertilization with 136 kg added N, respectively. 238 genotypes were sown in augmented design, where some genotypes are replicated and the others are not. Images at different field rows were taken randomly between 7:30 and 11:00 AM. "TensorFlow_preprocessing.ipynb" notebook can be used for data preprocessing and generation of additional 1200- 3600 images in processes of augmentation and segmentation.

This Nuts Classification Dataset is developed to support health-related research, particularly for identifying allergenic nuts through computer vision techniques. Given the health benefits of nuts, there is also a need to detect varieties that may cause allergic reactions. Based on online research, four types of nuts commonly associated with allergies were selected: Almonds, Cashews, Peanuts, and Pistachios. Nuts Samples were collected from local markets in Kaliakoir, Gazipur, Bangladesh, between November 16 and November 18, 2023. Images of these nuts were captured using a Xiaomi Redmi Note 9S smartphone. Originally in high dimensions (300x300 px), images were resized to 800x600 px to optimize file size while preserving quality. The dataset contains 4,390 images: 1,039 Almonds, 1,070 Cashews, 1,153 Peanuts, and 1,128 Pistachios. Backgrounds were removed to improve model learning efficiency; mages were taken with a white paper background and flash lighting for clarity and consistency. This balanced dataset, with backgrounds removed, is useful for computer vision projects to identify allergenic nuts using machine learning and deep learning models. It will also support health-related research by helping to classify nuts that may cause allergies. Original Data: Total: 4390 Images Data Type: JPG Dimension: 800 X 600 Without Background Data: Total: 4390 Images Data Type: JPG Dimension: 800 X 600