<![CDATA[The ability to recognize colors is an important aspect of human interaction with the environment, but it poses a significant challenge for people with visual impairments. This study aims to design and implement a prototype portable device called Smart Color Assistant to help people with visual impairments recognize colors. The system was developed based on an ESP32 microcontroller integrated with a TCS34725 color sensor for RGB data acquisition and a DFPlayer Mini module for sound output. The KNN (K-Nearest Neighbor) algorithm with a K value of 5 was applied to classify 16 common colors. The research approach applied was the Research and Development (R&D) method, which included several stages, namely needs analysis, system design, hardware and software implementation, and testing and evaluation processes. Based on the test results, it was found that the TCS34725 sensor had an average reading error of 2.61%. The system achieved a classification accuracy of 93.75% (15 out of 16 colors were detected correctly), with an average response time from detection to sound output of 850 milliseconds. Functionality testing of the buttons and sound output also showed performance in line with the design. Testing with visually impaired individuals confirmed the benefits of the device, although improvements are needed in terms of ergonomics and ease of use. It is concluded that this prototype has the potential to be an effective, portable, and responsive assistive device.]]>
Copyrights © 2026
