Thermometer Usage Correctness Classification Image Dataset

The current medical industry faces challenges in temperature monitoring due to insufficient accuracy and low data processing efficiency. During the traditional use of thermometers, users often cause data inaccuracies due to improper operation, and existing solutions mostly rely on manual reviews, which are inefficient and prone to errors. To enhance the reliability and accuracy of temperature monitoring, this dataset aims to automatically identify the use of thermometers through image classification technology and provide corresponding feedback. The dataset includes image data of users' correct and incorrect use of thermometers in different environments, collected using standardized shooting equipment to ensure data authenticity and diversity. During data processing, we implemented multiple rounds of annotation and consistency checks to ensure data quality. The data is stored in JPEG format, organized chronologically and by usage, to facilitate subsequent analysis and model training. The advantage of this dataset lies in its high-quality image data, with annotation accuracy reaching over 95%, and by introducing automated annotation technology, annotation efficiency has increased by 50%. Additionally, by employing diverse data augmentation methods, we have improved the model's generalization ability and solved the problem of poor performance of traditional methods in different environments, achieving a significant increase in model accuracy to over 90%.