Steps like feature extraction and categorization are crucial for identifying surface defects in steel. This study formulates and applies a multiclass support vector machine classifier as well as multiple statistical characteristics. First, four categories of statistical characteristics are suggested for various defect region properties. These image characteristics are used to construct a feature vector, which is then used to prepare a dataset of steel surface defects. The MSVM approach, which is based on feature matrices, is used for dataset testing and training. On the one hand, it can address problems with multiclass classification. On the other hand, it has a high classification efficiency and noise reduction capabilities. Finally, six different types of steel surface faults are discovered with a multi-class support vector machine classifier and multi-type statistical features. The experimental findings demonstrate the flawless accuracy and efficiency of our suggested multi-class classifier. Additionally, multi-type statistical characteristics support better classification performance.