Seismic data are consists of several amplitude values from high to low. Today’s seismic data is still a lot in the form of images, it is because the old measurement technique. So to use this seismic data again it must be digitizing. One of the techniques that can digitize image in computation field is image segmentation using K Means Clustering. So, first I choose the k or number of cluster based on the pixel color of the image. In seismic image, usually it consists of two colors, red and blue. Rather the red as the high amplitude or the low amplitude it depends on convention. In this case we use K = 6 to cluster the color on this seismic image. The result can be shown on the picture below.

Skills: Python, Machine Learning, Numpy, OpenCV, Scikit-Learn, Matplotlib