In response to the severe challenges of high myopia rates and prevalent poor learning postures among adolescents in China, this paper designs and implements an integrated solution incorporating intelligent visual protection, posture correction, and health management functionalities based on YOLOv8 and multi-modal sensing technologies. The system adopts a “device-edge-cloud” collaborative architecture, with the high-performance STM32F407VET6 microcontroller as the core processing unit, integrating computer vision and multi-modal sensing technologies. At the edge computing node, an innovatively deployed lightweight-optimized YOLOv8n-pose deep learning model achieves real-time high-precision recognition of poor sitting postures such as slouching, tilting the head, and hand-supported head postures, with an accuracy rate of 92.3%. Simultaneously, the system incorporates various sensors including the VL53L0X laser ranging sensor and BH1750 ambient light sensor to accurately monitor eye-screen distance, head posture, and ambient light intensity. Based on comprehensive analysis of multi-source data, the system implements multi-level, multi-modal active interventions through technologies such as TTS speech synthesis. All data is uploaded to the Huawei Cloud IoT platform via the ESP8266 Wi-Fi module using the MQTT protocol, supporting remote monitoring, big data analysis, and personalized health report generation. After a four-week field test and practical validation involving over 200 users, the system has demonstrated significant improvements in users’ sitting posture habits, showing positive effects in reducing myopia incidence and preventing health issues such as scoliosis. This provides a feasible technical pathway and practical reference for building an intelligent and personalized health-oriented learning ecosystem for adolescents.