My main research direction is automatic stratigraphic division, mainly designing deep learning algorithms as well as image processing techniques, aiming to improve the degree of automation and accuracy of stratigraphic division, to solve the problems of much manual intervention and low efficiency in traditional methods, and to promote the development of geologic analysis towards intelligence and efficiency. In my research, I will use a variety of deep learning frameworks, such as TensorFlow and PyTorch, to build and train neural network models. These models are able to learn features of the formation, including texture, color, structure, and morphology, from a large number of geological images. I also focus on how to improve the generalization ability of the algorithms so that they can work stably in different geological environments. In addition, I investigate how to combine traditional image processing techniques, such as edge detection, feature extraction, and image segmentation, to enhance the performance of deep learning models. These techniques can help the models better understand the geological structures in the images and improve the accuracy of automatic division.