The invention relates to a blasting three-dimensional point cloud data processing technology, in particular to an automatic identification method for blasting incomplete holes. The method comprises astep 1: performing cloud point original data processing; a step 2 of calculating the curvature features of pre-processed point cloud data and extracting the curvature feature points; a step 3 of usingthe curvature feature points to fit the space line, and identifying the incomplete holes based on the space line. The method realizes automatic, high-speed, high-efficiency and high-precision identification of blasting incomplete holes by extracting the curvature feature points of the high-precision three-dimensional point cloud data to fit the incomplete hole position axes, and solves the defects that the traditional modes are incomprehensive in incomplete hole identification range, difficult in identification, low in efficiency, and poor in consistency of identification standards. The invention provides an automatic blasting incomplete hole identification method that is suitable for future intelligent and refined blasting requirements. The invention also provides an automatic extractionmethod for blasting incomplete hole features on the basis of the above automatic identification method for blasting incomplete holes.