With the ever-increasing demand for logistics, improving the efficiency of work in logistics warehouses is an important issue. Research is being conducted to improve the efficiency of picking work by classifying picking methods according to the differences in the size of demand for products. There are two types of picking method: category-picking, where products are picked in bulk, and multi-picking, where orders are batched. Previous research has not sufficiently considered the location of temporary storage areas used when picking category products, or the capacity of pickers. In addition, the classification method for optimal picking patterns in previous research had the problem of eliminating some of the combinations of products to be picked from the start. This paper clarifies the method for calculating the number of round trips to be considered when classifying picking patterns and enables the determination of efficiency picking patterns using dynamic programming. The method proposed in this paper can encompass previous research and provides new insights into the classification of realistic picking patterns.