In the field of infrared quantitative defect detection, the current theoretical methods simulate the defects mostly by using a rule model and then obtain the size of the detected defect by using an image comparison method or the first and second derivative of the image. All of these methods may bring about a large data error. The heat conduction among particles based on the particle theory is applied in quantitative defect detection for the first time. According to a simulated defect, the heat conduction equation is established. It provides a new idea for the quantitative infrared defect detection.