Microwave, as a clean and efficient energy, can solve growing levels of pollution and energy problems. But In the process of industrial production, the use of high-power microwave will cause hot spots, thermal runaway, or even explode. Therefore we need to achieve intelligent control of microwave, to ensure the safe and efficient use of microwave energy. Accurate temperature information is the basis for intelligent control, and there are many methods to monitor temperature in microwave oven, for example, optical fiber and infrared temperature measurement and ultrasonic thermometry. But these methods suffer from high attrition rate locality, or difficulty in installation problems. To solve the problems, we propose a temperature monitoring approach using image processing and colorimetric technique. To evaluate the proposed approach, we established a microwave heating system with camera and fiber, developed temperature measurement software to monitor temperature, and visualized the temperature field for the heating object in the microwave oven. The experiments show that the proposed approach is non-contact with low cost, easy to installation, and accurate on the reconstruction of temperature field.