Eastern Chinese landscape paintings and Western Impressionist landscape paintings take natural scenery as their creative theme, showcasing their unique artistic value by presenting landscape spatial composition, aesthetic experience, and cultural transmission. The attention mechanism, originating from human visual research, can identify the focus areas the human eye concentrates on within images. It has now been widely applied in deep learning and has achieved remarkable results. This study combines science and art by examining Shanshui and Impressionist landscape paintings, collecting more than 4,000 high-definition art electronic painting images, and using the attention mechanism method to explore the differences and similarities in visual representation and aesthetic mood between the two. The results reveal that Eastern and Western paintings share unique composition, colour, spatial hierarchy, and visual focus characteristics. Researchers have successfully applied the attention mechanism to painting recognition and landscape design due to its simple, high-speed, fully-automatic, and accurate features. Leveraging the advantages of Eastern and Western artistic styles and integrating them allows for unearthing potential traditional aesthetic principles in modern design and constructing a contemporary landscape composition model from three aspects: space composition, visual characteristics and place objects. From the perspective of the objectivity of big data and the universality of the public perspective, this landscape composition model globally blends Chinese and Western artistic features, which is expected to provide a more prosperous and diverse array of practical ideas for contemporary cultural landscape design, fulfilling the modern human desire and pursuing a better living environment.
Su Yaohui is a PhD student in Environmental Art Design at the Academy of Arts & Design, Tsinghua University. She received her Master’s degree in Landscape Architecture from Peking University, and her research interests are cultural landscapes.