Abstract: Aiming at the problem that the traditional single characteristic smoke monitoring method is easily disturbed in the coal mine environment, this paper analyzes the advantages and limitations of applying new multi-feature smoke detection technology in coal mine scenarios. We propose that using a dual-wavelength photoelectric smoke detection technology for monitoring coal mine fire smoke and test its performance and issues in coal mine scenarios. Through designing targeted data processing algorithms and introducing an auxiliary monitoring method for carbon monoxide concentration, a dual-wavelength photoelectric smoke-based multi-feature smoke monitoring method for coal mine fires is developed. A technical prototype based on dual-wavelength smoke detection is designed, tested, and its performance is demonstrated. According to experimental verification, this technical prototype has better anti-interference capabilities than traditional prototypes against coal and dust, water vapor, and wind flow in coal mine environment.