Patient Basic Care Monitoring Dataset

The current medical industry faces challenges in basic care monitoring, such as incomplete patient data acquisition and heavy workload on nursing staff. Existing monitoring systems often provide limited image data with a lack of high-quality annotation information, restricting the application of technology. This dataset aims to provide high-quality patient basic care monitoring images to help solve issues of insufficient image data and inaccurate annotations. In the construction of this dataset, professional medical equipment is used to capture images in a real medical environment, and quality control is conducted through multiple rounds of annotation and expert review to ensure accuracy and consistency of the data. The data is stored in JPG format, organized by patient ID for ease of use and management. The annotation accuracy of the dataset reaches over 95%, and the data consistency has been rigorously tested to effectively support the needs of object detection tasks. Additionally, new data augmentation techniques have significantly enhanced the diversity and richness of the dataset, thereby improving model performance in practical applications, with prediction accuracy increased by 15%.