Green eco-housing architecture has attracted much attention in today’s environmental context of rapid socio-economic development and continuous improvement of people’s quality of life. This study takes design ethics as the theoretical basis and combines the requirements of ecological responsibility in housing design to design a two-dimensional image of sustainable housing. The Canny operator is utilized to detect the edges of the 2D image and extract its contour, the VGG-16 network is used to extract the main features in the 2D image to generate feature vectors, and then the convolutional layer generates the 3D design model of sustainable housing. Subsequently, the SuperMap platform is used as a vehicle to design a sustainable housing design system to realize the ethical issues and ecological responsibility. It is verified that the 3D modeling method proposed in this paper is able to accurately align the sustainable housing design images designed by modeling. It is also found that the building density and green space ratio in the sustainable housing designed based on the methodology of this paper are 21.85% and 42.59%, respectively, which are able to realize the ethical issues and ecological responsibility in sustainable housing design. This paper provides a reference method for the effective design of sustainable housing and lays a strong foundation for the conservation and sustainable development of the earth’s ecology.
