Lead apron detection using convolution neural network

Lead aprons are essential for personal protective equipment used in laboratory environments to protect workers from harmful radiation exposure. A lead apron contains materials meant to attenuate radiation by scattering and absorption during safety involving diagnostic and therapeutic radiation procedures. However, ensuring consistent lead apron usage presents challenges because many facilities rely on manual observation and self-reporting methods especially in busy environments where continuous monitoring is difficult to maintain. Manual monitoring often fails to provide coverage and documentation. Therefore, the convolutional neural network method is used to detect lead apron and provide efficient monitoring systems that help to ensure with radiation protection. This research uses a convolutional neural network model approach designed for small dataset scenarios. Lead apron detection involves binary classification with two categories such as with apron and without apron scenarios. The dataset comprises 295 grayscale images processed at 150 x 150 pixel resolution and annotated for binary classification. From the test result, the model achieved performance with a precision value of 97 %, recall value of 97 %, and F1-score value of 97 %. The model achieving good 95.7 % for lead apron detection and precision score 98.5 % for without apron detection scenarios.