Research On Fashion Colors Of Clothing Based On Deep Learning And Color Feature

The fashion industry is constantly changing,and fashion designers need to keep up with the latest trends and create exciting new designs.Analyzing clothing color trends can provide insights into the latest trends and help designers create fashionable clothing.However,manually analyzing clothing color trends is both time-consuming and labor-intensive.Therefore,an automated method for analyzing clothing color trends is needed.The existing research methods are not ideal in terms of real-time and authoritative data,and the credibility of the conclusions obtained is not high.Therefore,this article focuses on the real-time and authoritative nature of research data,improving the accuracy of popular color research.The main contributions are as follows:(1)This article studies fashion popular colors based on e-commerce platforms and proposes using an improved Mobile Net V3 model to process clothing classification problems faster.Using temporal e-commerce platform sales data as samples,Grab Cut algorithm is used to analyze the main colors of clothing.KMeans algorithm is then used to calculate the proportion of main colors and mixed colors.Finally,time dimension,clothing type dimension,and brand dimension are analyzed for the main colors of clothing,Obtain trend data on fashion colors.Research has found that clothing popular colors based on improved Mobile Net V3 have higher real-time data,larger data capacity,and faster analysis speed compared to current research methods.(2)A research process for clothing popular colors in monitoring scenarios was proposed,which combines YOLOv5 object detection algorithm and Grab Cut image segmentation algorithm for clothing popular color analysis.The main steps include data collection,object detection,image segmentation,and popular color analysis.In terms of data collection,this article uses the Deep Fashion2 dataset for training and testing the model.In terms of object detection,this article uses the trained YOLOv5 model to detect clothing objects in the input image and extract the corresponding ROI(region of interest).In terms of image segmentation,this article uses the Grab Cut algorithm to segment the target area in order to separate clothing objects and backgrounds.Finally,in terms of popular color analysis,the segmented clothing objects are extracted and analyzed to obtain the popular color of the clothing.Overall,this method combines the YOLOv5 object detection algorithm and Grab Cut image segmentation algorithm to quickly and accurately analyze the popular colors of clothing,improving the efficiency and accuracy of the entire process.It is of great significance for clothing design and marketing.By analyzing massive clothing product data,it can more accurately grasp market demand and consumer preferences,and apply the research results of this article to industrial production,It can provide better products and services for clothing brands and businesses,and also contribute to public safety.