Given the two-dimensional bin packing problem under the guillotine constraint, an improved priority heuristic (IPH) algorithm with rapid solving ability is proposed. IPH can directly divide and fill the residual space without any iterative searching. For this reason, the priority placement method in PH algorithm is developed, and an improved bricklaying space division rule based on maximization of newly-formed large-space area and minimization of small-space area is designed. Based on the heuristic strategy that can obtain a better solution with higher probability, the rule designed is favorable for placement of large rectangles and less likely to form wasted space. Comparative experiments with standard datasets shows that IPH can effectively solve large-scale cases within very short time and produce the optimal filling effects of many cases for the first time. The new algorithm achieves the 2D-GRPP optimization allocation under the direct operation mode and offers some technical reference for relevant research on GRPP.