A noncontact and high-precision detection method based on laser ranging sensing is investigated for measuring the surface profiles of stainless steel lap laser weld joints. The influences of roughness and inclination of the workpiece surface on detection results are also studied. Moving-mean filtering and image tilting correction methods are adopted, and a detection data filtering and correction algorithm is developed. Based on feature point recognition of the detected profile curves of laser weld joints, a method for detecting and calculating the residual and collapse heights of laser welding joints is established. Results show that the profile curves of laser welding joints obtained by these methods are in good agreement with the corresponding metallographic detection results; additionally, weld-surface conditions are well reflected