Aiming at the current low efficiency and insufficient personalization of 3D space printing, a fast 3D object printing method based on the subdivision of the region of interest is proposed. First, use the region of interest algorithm to extract the most significant part of the color image, and then perform secondary image processing through a new gray-scale image re-segmentation algorithm to obtain images with different saliency region levels, the regional graded images and the original images are merged to obtain a color image with regional grading; then through a three-dimensional texture mapping method, the resulting color image is mapped to the target three-dimensional model, and the surface of the three-dimensional model has the required texture information. A new multi-nozzle control technology is used to rationally arrange different nozzles to handle different areas. Finally, through signal transmission, the print file is transmitted to the robotic arm, and the robotic arm performs work printing. The three-dimensional inkjet method has a total working time of 100.34 s for a sphere with a radius of 100 mm, which has high working efficiency. Compared with traditional 3D printing and printing based on traditional regions of interest, the proposed re-segmentation algorithm based on regions of interest significantly improves the printing efficiency while ensuring the accuracy of printing, and is generally better than the above algorithms in terms of performance indicators.