Research on Intelligent Fractal Lithological Feature Recognition Based on Big Data

To address the issues of low accuracy, poor applicability, and slow identification speed in current lithological feature recognition methods, a new lithological feature recognition method based on fuzzy clustering algorithms is proposed. First, use the fractal signal algorithm to traverse each pixel in the hyperspectral image, obtaining the fractal signal curve and fractal signal image at a specific scale for the hyperspectral curve. Then, extract the lithological characteristics of the rocks based on the resulting images. Next, obtain the correlation between measurement parameters of lithology features to be identified through cosine angles. Classify rocks using a fuzzy similarity matrix, calculate the maximum membership degree of lithology patterns, and then achieve intelligent fractal recognition of lithology features based on the membership degree values. The experimental results indicate that this method can quickly and accurately identify the lithological characteristics of the test samples, with strong applicability and suitability for fractal identification of rock types.