In order to improve the planning efficiency of urban landscape, this paper proposes a combination design method of urban landscape construction based on grid division and a spatial optimization model of urban landscape based on particle swarm algorithm to optimize the spatial and pathway layout of urban landscape that takes both economy and ecology into account. The original landscape image was mapped with 3D remote sensing image to generate a 3D image model, and the gradient decomposition method was used for image sampling. Then the multi-dimensional dynamic feature distribution model of urban landscape was constructed, on which the urban landscape area grid was divided to realize the landscape construction combination design. Using particle position to simulate the meta-space layout results of landscape type raster images, the optimization of landscape pattern space and path is completed. The experiment proves that the algorithm in this paper reduces the influence of multiple types of perturbations on the landscape layout results, and the spatial optimization model of urban landscape pattern based on particle swarm algorithm realizes the organic coupling of quantitative and spatial optimization, which not only improves the utilization rate of the urban land, but also substantially reduces the risk index of the urban landscape, and meets the design expectations.