Aiming at the positioning problem in the stitching construction of tunnel boring machine（TBM） steel arches,a detection method for steel arches hinged hole location based on machine vision was proposed.According to the scene environment of steel arches splicing,the image acquisition system was designed;normalized cross-correlation（NCC） template matching algorithm with rotation invariance was used to quickly match the region of interest where the steel arch hinged hole was located;based on the gray gradient on the one-dimensional pixel sequence,the edge points of the outer contour of the hinged holes were determined,and the center of the hinge holes of the steel arch was determined by using the ellipse fitting method based on random sample consensus（RANSAC ）algorithm,so the positioning of the steel arches was realized,which provided the position and pose information for the steel arches splicing.The results show that by using this method,the high efficiency and high precision positioning and detection of steel arch hinged holes are realized.