Urine color is an indicator of health status, especially for patients undergoing medical intervention treatment of urinary catheterization. Urinary tract infections are the most frequently developed infections among patients receiving treatment in medical institutions. Such infections can be detected from urine color, like in the case of the purple urine bag syndrome. However, it is a di±cult task for non-nursing care sta® and even the nursing sta® to correctly conduct naked-eye identi¯cation without proper tools. To better assist both nursing and non-nursing care sta® with the detection of infection signs in urine bag patients, a urine color automatic identi¯cation device has been developed. The device is based on microcontroller framework and color quantization algorithm. A hybrid color quantization algorithm and two features were proposed to identify the urine color. The identi¯ed color, as query data instead of human-described color keyword, can be used to retrieve the information from the database and then¯nd possible symptoms for early warning. Instead of the nursing sta®, the device can automatically identify the patient's urine color. From experimental results, the device with the proposed algorithm shows its capability and feasibility of the urine color automatic identi¯cation.