The work aims to conduct an in-depth study of design input to deal with the random design output generated when using the Midjourney auxiliary design tool, thereby enhancing the scientificity and credibility of the first step of intelligent auxiliary design. With the design practice of home grown products as an example, users' perceptual needs and functional needs were summarized and quantified through Kansei Engineering (KE) and Analytical Hierarchy Process (AHP), and a systematic and scientific research framework was built for "design input". Finally, its feasibility and effectiveness were verified. Through the application of KE and AHP methods, the randomness of drawings generated using Midjourney were successfully reduced, and the overall design output was closer to the user needs. At the same time, it also exposed problems such as inappropriate handling of design details by intelligent design tools. The KE and AHP methods standardize the design input of Midjourney tools, improve the controllability and rationality of design output, and also provide a more solid theoretical foundation and scientific basis for the first step of applying Midjourney to design output. At the same time, the emergence of artificial intelligence design tools also puts forward new requirements and suggestions for the future development and positioning of designers.