To address the problems about the difficulty in accurate recognition of distribution features of gas flow center at blast furnace throat and determine the relationship between gas flow center distribution and gas utilization rate, a method for recognizing distribution features of blast furnace gas flow center was proposed based on infrared image processing, and distribution features of blast furnace gas flow center and corresponding gas utilization rates were categorized by using fuzzy C-means clustering and statistical methods. A concept of gas flow center offset was introduced. The results showed that, when the percentage of gas flow center without offset exceeded 85%, the average blast furnace gas utilization rate was as high as 41%; when the percentage of gas flow center without offset exceeded 50%, the gas utilization rate was primarily the center gas utilization rate, and exhibited a positive correlation with no center offset degree; when the percentage of gas flow center without offset was below 50% but the sum of the percentage of gas flow center without offset and that of gas flow center with small offset exceeded 86%, the gas utilization rate depended on both the center and the edges, and was primarily the edge gas utilization rate. The method proposed was able to accurately and effectively recognize gas flow center distribution state and the relationship between it and gas utilization rate, providing evidence in favor of on-line blast furnace control.