This paper proposes coupled hidden Markov models (CHMM) for analysis of steel surfaces containing three-dimensional flaws. Due to scale on the surface, the reflection property across the intact surface changes and intensity imaging fails. Hence, the light sectioning method is used to acquire the surface range data. The steel block is vibrating on the conveyor during data acquisition which complicates the task. After depth map recovery and feature extraction, segments of the surface are classified by means of CHMMs. We present classification results of the CHMM and compare them to the naive Bayes classifier. The CHMM outperforms the naive Bayes approach.