Welcome to Hydra Research Lab, Smart Digital Innovation Lab

Dataset / Cavendish Banana

Cavendish Banana Ripening Dataset

Bananas are rich in minerals and various nutrients. that the body should receive and provide up to 100 kcal of energy per unit. Because in bananas there are 3 types of sugar: sucrose, fructose, glucose, and dietary fiber. Thus our body will get energy immediately. Two bananas can provide energy for up to 80-100 minutes (160-200Kcal) and also have other benefits such as stress relief, brain power, reduce the occurrence of ulcers in the stomach, reduce anemia reduction of the occurrence of cramps, nourishing the nervous system, etc.

Our team built the Cavendish banana dataset while ripening. The dataset aim to (1) Banana Segmentation from background using DIP (2) DeepSemeticInstanceSegmentation (3) Ripeness Level Prediction for both DIP method and CNN Classification. We already labeled (annotation) for the data. The ripeness prediction level divied into 8 level.

To Download

Please read and accept the Agreement and Disclaimer below before download. If you download this dataset, we count you already accepted the agreement and disclaimer.

1. The dataset must be only used for non-commercial or educational purposes.

2. No responsibility or any guarantee for any incidents, or damages caused by the direct or indirect usage of this dataset.

3. This dataset is copyright of Rajamangala University of Technology Srivijaya,for commercial usage please contact Technology Licensing Office.

4. If you used the dataset, please consider to referencing paper [พงศกร เจริญเนตรกุล, นงนาฎ ระวังวงศ์, ชัยสิทธิ์ ชูสงค์, อรรถพล คงหวาน, กิติศักดิ์ วัฒนกุล,ณัฐพล หนูฤทธิ Ripeness Level Classification of Cavendish Banana Using Digital Image Processing and Deep Learning การจำแนกระดับความสุกของกล้วยหอมอัตโนมัติด้วยการประมวลผลภาพดิจิทัลและการเรียนรู้เชิงลึก, ETCI-CARD-2023, ครั้งที่ 15, pp. 31 - 35, 2023].

Gallery

Dataset / Medical Waste Dataset

Medical Waste Dataset

Our team built Medical Dataset 41 Classes from real captured wastes/equipments. The images/video motions are keep in different environments e.g. Controlled Environment Box, Controlled Light Indoor, Dark Zone Indoor, Different Background Indoor, Different Background Outdoor, Overlay Items Indoor (Simulate to ConveyorBelt). All 3,868 cropped images are used for deep learning models training and 7,140 cropped images used for testing the model.

There are 41 classes for medical waste included : 1 Way Connector for Foley,2 Ways Connector for Foley,2 Ways Foley Catheter,3 Ways Connector for Foley,3 Ways stopcock,Alcohol Bottle,Alcohol Pad, Boot Cover,Cotton Ball,Cotton Swap,Dilator,Disposable Infusion Set,Extension Tube,Face Shield,Front Load Syringe,Gauze Pad,Glove,Guide Wire,Liquid Bottle,Mask,Nasal Cannula,Needle,NG Tube,Oxygen Mask,Pharmaceutical Product,Pill,Pill Bottle,PPE Suit,Prefilled Humidifier,Pressure Connecting Tube,Reusable Humidifier,Sodium Chloride Bag,Sterile Humidifier Adapter,Surgical Blade,Surgical Cap,Surgical Suit,Syringe,Tracheal Tube,Urine Bag,Vaccine Bottle,Winged Infusion Set.

To Download

Please read and accept the Agreement and Disclaimer below before download. If you download this dataset, we count you already accepted the agreement and disclaimer.

1. The dataset must be only used for non-commercial or educational purposes.

2. No responsibility or any guarantee for any incidents, or damages caused by the direct or indirect usage of this dataset.

3. This dataset is copyright of Rajamangala University of Technology Srivijaya,for commercial usage please contact Technology Licensing Office.

4. If you used the dataset, please consider to referencing paper [พงศกร เจริญเนตรกุล, กิติศักดิ์ วัฒนกุล, ชัยสิทธิ์ ชูสงค์, นงนาฎ ระวังวงศ์, อรรถพล คงหวาน, ณัฐพล หนูฤทธิ "การแยกประเภทภาพขยะทางการแพทย์ด้วยการเรียนรู้เชิงลึก Medical Waste Image Classification Using Deep Learning",THE 46TH ELECTRICAL ENGINEERING CONFERENCE (EECON-46), pp. 312 - 315, 2566.].

Prediction Demo

Gallery

41 Classes of Medical Waste in the dataset

Sample Images

Dataset / Topview Depth Human Action

Dataset / Recycle Waste Dataset

Waiting for Labeling 20/100%