The work aims to propose the stochastically seeded clustered halftone algorithm with respect to the problem that the binary image obtained by error diffusion halftone algorithm has obvious structured artifacts. The stochastically seeded dots were generated by means of blue noise halftone threshold matrix algorithm. The Delaunay triangulation network was constructed with the triangulation network growth algorithm. After the pixel members of triangulation network were confirmed, based on the position of pixel points distributed in the triangular area, the threshold of pixels at such position was obtained. Finally, the halftone images were obtained after the comparison on the thresholds of continuous-tone image and the threshold matrix. Compared with the error diffusion algorithm, the proposed method could reduce the artificial texture of halftone images, so that the images had better visual effects. The proposed algorithm can improve the quality of the halftone images, and the experimental results illustrate the effectiveness of such method.