<![CDATA[With the limited number of nurses compared to the number of patients in a health center, a monitoring system is needed. A computer vision-based patient bed monitoring system was created to assist nurses in supervising patients so as to minimize the risk of falling out of bed. The Haar cascade method is used as a method in computer vision to identify the sleep position of the patient captured by camera. The Haar cascade classifier is trained to recognize the patient’s sleeping position so that it can accurately distinguish between patients sleeping in the middle of the bed and patients sleeping on the edge of the bed. The test results show good performance of the system because the system can recognize the patient if the patient is in bed and can recognize the patient’s location in bed whether it is in the middle or at the edge of the bed. When the system recognizes the patient is in a position at risk of falling, namely on the edge of the bed, the system sends a warning message to the nurse and the nurse can determine the next action.]]>
