Automating the industrial quality control of natural wooden surfaces is challenging. The core problem in this domain of computer vision is that each surface is composed of statistical features, making each one unique and difficult to analyze due to natural variability. Hence, locating defects and defining the normality of such surfaces for anomaly detection is a difficult task. This is further complicated by the subjective judgement of inspectors during manual inspection of wood surfaces, leading to inconsistency in deciding whether surfaces with certain defects should be kept or discarded. To address these challenges, we first present a method to capture image data in real-time in a production setting, as well as a graphical user interface that allows operators to interact with the system. Besides the visualization of captured data and detection results, this interface allows operators to provide feedback for further refinement of the system. Furthermore, we propose several AI Methods that were trained and evaluated using data sets created from the captured image data. These approaches combine methods ranging from the detection of anomalies and the recognition and classification of objects to the similarity analysis of different wood surfaces. Additionally, this paper provides valuable insights into the effectiveness of state-of-the-art methodologies and the creation of annotated data sets in this domain.