The uniformity of the asphalt mixture is severely affected by aggregate segregation during the paving process, and there are many studies that indicate that segregation may lead to early pavement damage. This study aims to develop a real-time asphalt mixture segregation identification algorithm during paving process based on digital imaging technique and the static moment method. Static moments of the aggregate particles relative to the four boundaries of the image were calculated as an evaluation parameter, and the dispersion coefficient between the static moment of segregated asphalt mixture and the value of uniform asphalt mixture was calculated to quantify the degree of segregation. The algorithm was applied in a filed construction project in Yunnan, China. Comparing the result from the proposed algorithm with the sand patch test shows that the proposed algorithm is capable of detecting the segregation of an asphalt mixture in real-time and that it is accurate in identifying the degree of segregation. In addition, the proposed algorithm is more accurate than the sand patch test for areas where the gradation meets the standard but the distribution of the aggregate particles is not uniform.