Abstract: Accurate emissivity parameter settings are crucial for infrared temperature measurement and for infrared image recognition of coal and gangue. Aiming at the accurate application of infrared thermal imager in the “liquid intervention + infrared detection” coal/gangue recognition method, before the infrared thermal imager detects coal and gangue, its own emissivity needs to be consistent with the emissivity of coal and gangue to ensure the accuracy of the test results. Based on the matching method of infrared thermal imager and surface thermocouple, this article builds an experimental platform and device for measuring the emissivity of coal and gangue by matching method, and carries out the extraction experiment of coal and gangue emissivity features under different temperature conditions. The experimental results show that the surface roughness of coal and gangue under different temperature conditions constitutes an internal factor influencing the emissivity difference of coal and gangue itself. The emissivity of coal and gangue decreases with the increase of sample temperature,following a power function. The relationship between emissivity and temperature of different kinds of coal and gangue is presented, and the emissivity value at room temperature is predicted. The accuracy of the test results of the matching method is verified through the application of the inverse method. It is observed that the error value of the measured value decreases with the increase of emissivity. These research findings can offer valuable support for coal/gangue recognition technology based on infrared imaging.