Intelligent Design of Agricultural Product Packaging Layout Based on Reinforcement Learning

AbstractAlong with the rapid development of e-commerce logistics, the volume of business in online sales of agricultural products is also growing, but e-commerce for agricultural products also suffers from a lack of funding for packaging design and difficulties in matching the packaging image with the product itself. One of the solutions to these problems is to create an intelligent design platform through a machine learning paradigm that can quickly and inexpensively meet the personalised design needs of users. One of the key technologies is intelligent layout design. In this paper, a deep deterministic strategic gradient (DDPG) algorithm is used to output the layout of the packaging layout by taking the graphic design elements of the packaging as input data, while taking into account constraints such as aesthetic principles, to achieve automatic layout and automatic output of the layout of the packaging appearance. Finally, through a user survey, it is found that the packaging layout designed using the algorithms in this paper can basically meet the requirements.KeywordsPackaging of agricultural productsAutomatic layoutReinforcement learning