An evaluation method for the segmentation of ferrograph image based on grey relational analysis

The segmentation of ferrograph image is one of the important steps of the intelligent ferrography analysis, which will influences the results of the subsequent wear particle recognition and fault diagnosis. In order to improve the accuracy of particle segmentation, a new evaluation method on the segmentation quality of ferrograph image is put forward based on grey relational analysis. This method evaluates the qualities of segmentation results of wear particles using the color ratio parameters in RGB and CIELAB color space. The experimental results show that this method can automatically evaluate segmentation quality with high precision.