With the ongoing advancement of artificial intelligence, which is progressively transforming technical paradigms within the electrical engineering domain, this research concentrates on the application of intelligent perception and electromagnetic detection technologies. The objective is to design and develop a sophisticated intelligent wardrobe system. Utilizing the STM32F103 microcontroller as its computational and control core, the system implements algorithms including linear regression models and finite state machines. It further integrates a multi-sensor fusion-based perception module, comprised of temperature/humidity sensors, photoresistors, and smoke sensors, coupled with Wi-Fi-based electromagnetic detection technology. This integrated approach tackles the limitations inherent in traditional wardrobes—primarily their unitary functionality and absence of intelligent awareness—as well as the deficiencies of current intelligent products. These often rely solely on preset thresholds, lacking dynamic, algorithm-driven intelligent decision-making, which consequently results in delayed responses, inefficient energy use, and elevated operational costs. The proposed design successfully achieves multiple functionalities, encompassing predictive assessment of mildew risk, automated disinfection, and active dehumidification control.