At present, there is a lack of effective methods for real-time detection of welding profile feature parameters in laser welding processes. In this paper, a method for detecting the appearance of weld seams based on optical low-coherence technology is proposed. Firstly, the three-dimensional point cloud information of the welding surface is obtained in real time using the fast measurement speed of optical low- coherence technology. Secondly, the three-dimensional contour of the weld seam point cloud is extracted, and the contour line and curve fitting are realized by an improved RANSAC algorithm and gradient descent algorithm, respectively. Finally, the feature equation is used to determine the feature points of the profile, and the feature parameters of the weld seam profile such as weld height and width are obtained. Experimental results show that the proposed method has both real-time and accuracy advantages, with a detection accuracy of up to O.004mm, satisfying industrial production requirements.