The invention discloses a method for monitoring the gas flow development process in the blast furnace material distribution period and predicting the gas utilization rate. The method comprises the first step of data collecting and processing, the second step of infrared image processing, the third step of image feature extraction, the fourth step of material distribution period gas flow distribution dynamic changing model construction, the fifth step of gas flow center feature extraction, the sixth step of image and material surface position calibration, the seventh step of material distribution period gas flow center dynamic change model construction and the eighth step of gas utilization rate prediction model construction. According to the method, collected furnace roof infrared images are treated, the clustering algorithm, the statistics method, the feature recognition technology and the mode recognition technology are utilized for achieving dynamic tracking of the material distribution period gas flow distribution state and falling point distribution of the gas flow center on the material surface, the neural network algorithm is utilized for finding the relation between the material distribution period gas flow development process and the gas utilization rate, real-time prediction of the gas flow utilization rate is achieved, and assistance is provided for intelligent production.