The Application Of Study On Soft Computing Method For Product Image Design

In the 21 st century,the consumer-oriented becomes the mainstream,and in the field of product design,it becomes the key to meet the emotional needs of users.Based on this background,product image design has been paid more and more attention.The user's choice of product varies from person to person and has the complexity and uncertainty that can't be captured.Soft computing is a kind of new computer technology,which can simulate and learn human thinking mode in fuzzy or uncertain environment.Therefore,it is very suitable to apply soft computing to the research of consumers' cognitive mode of products.In order to design the products which can satisfy the user's inner emotional needs to the maximum extent,this research uses the soft computing method based on the image to carry on the following research:(1)Through the analysis of the relevant theories and methods of product image research,the representative experimental samples are extracted by using the concept of inductive engineering,and the proportion of target images is evaluated by entropy method,and the target images which affect the product modeling elements are obtained.(2)Frame glasses are selected as the experimental object of this study,and the form characteristics of the experimental samples are described in a qualitative way.Firstly,the technical methods to study the relationship between the design elements of product form and the image of the object are analyzed,and then,A method of combining grey relational analysis and entropy.(3)The technique of product image design was constructed based on the grey theory and neural network,particle swarm optimization algorithm is proposed to optimize the product image modeling design method of grey neural network.The shortcomings of GRA-GNNM model such as low precision,long time and different results of each training are made up.Finally,the result of optimization is shown by 3D modeling method.Through the questionnaire survey and statistical analysis of the optimized model samples,the feasibility of the product image design method proposed in this paper is proved.It can provide effective assistance and support for product image design.