Analysis of Rock Texture Feature Extraction Technology in Huangsongyu Geopark
Huangsongyu Geopark is located at the intersection of the southern foothills of Yanshan Mountain and the northern end of North China Plain.The dolomites,basal conglomerates,sandstones,shales,etc.exposed in the area are widely representative and typical in North China.It is particularly important to identify rocks'valuable features for accurately recognizing their types.By Matlab simulation platform,the research is conducted on some key technologies in rock feature extraction.Wavelet transform is used for filtering and denoising preprocessing,and line features are detected based on Hough transform.The two-dimensional image is dimensionally reduced and projected into one-dimensional curve signals in the vertical and horizontal directions.The texture structure features of the image are analyzed based on parameters such as mean,standard deviation,skewness,and kurtosis.Compared with Gaussian smoothing algorithm,the Mallat algorithm of wavelet transform can effectively achieve filtering and denoising preprocessing,and greatly improving the signal-to-noise ratio of the image.Compared with the Prewitt edge detection operator,using the Canny edge detection operator to extract binary images for Hough transform detection is more conducive to accurately extracting the linear texture information of rocks.Combined with one-dimensional projection curve signal features,it has strong recognition ability for linear textures.The use of correlation analysis to set different permission parameters and connect them in series to form collaborative data can achieve multi feature fusion and improve the expression ability of complex rock structures.
NETL
NSTL
万方数据
