Inoculation Powder Quantity Classification Dataset

The Inoculation Powder Quantity Classification Dataset (IPQCD) is collected to support the development of machine learning models focused on classifying images based on the amount of inoculant powder applied in the casting process. This dataset provides a valuable resource for training artificial intelligence models aimed at improving real-time monitoring systems in the casting industry.
The dataset consists of 6,908 images captured during the inoculant powder spraying process, with each image representing a specific quantity of inoculant powder. The dataset contains 20 distinct classes, each corresponding to a specific quantity of the inoculant powder being sprayed. The images in this dataset were captured under a variety of conditions to ensure robust model performance in real-world scenarios. By accurately classifying powder quantities, the models trained on this dataset can optimize the inoculant spraying process and help achieve consistent product quality.
For detailed information regarding the quantity of inoculant powder sprayed in each class and frame, please refer to the file readme.pdf.