3D surface form error evaluation using high definition metrology

This paper presents an approach to evaluate 3D surface form error of machined surface using high definition metrology that can measure millions of data points representing the entire surface. A data preprocessing method was developed to convert the mass data into a height-encoded and position-maintained gray image. With the converted image, a modified gray level co-occurrence matrix method was adopted to extract 3D surface form error characteristics, including entropy, contrast and correlation. Entropy measures the randomness of surface height distribution. Contrast indicates the degree of surface local deviations. Correlation could be used to identify different machining techniques. These characteristics can be used with flatness together to evaluate 3D surface form error of large complex surface.