泰国地区消费者脚型数据对应运动鞋尺码预测数据

中国制造的运动鞋正通过亚马逊、全球速卖通等平台销往全世界不同的国家，不同国家的消费者脚型对应合脚的运动鞋尺码存在差异。通过将三维足部扫描测量算法模型接入线上商城以采集泰国消费者的足长、足宽、足弓高和跖围等脚型数据，经过三维建模还原以及足鞋数据匹配，分析预测出最适合泰国消费者脚型的运动鞋尺码。帮助开展运动鞋出口销售业务的企业降低由于鞋不合脚导致的销售退换货率，同时减少退货换货过程的运费成本。1.数据采集：泰国消费者在买运动鞋前使用手机拍摄3张自己的足部脚型照片上传至系统；2.数据处理：根据泰国消费者上传的照片调用三维足部扫描测量算法模型，使用YOLO图像识别模型识别出消费者的脚型图像，再调用RCF边缘检测算法，从照片中完整提取消费者的完整脚型。根据用户填写的尺码数据，进行三维建模还原足部完整形态；在还原过程中可能出现图像缺口，按照尺码±0.5码处理，阈值限制为±1码，还原最小缺口的三维足部模型；3.数据分析与预测：根据足部三维建模结果，基于二分法规则等比例分析计算出消费者的实际足部尺寸数据，参考国际标准的足鞋尺码对应表预测出该消费者足部尺寸对应的最合脚的运动鞋尺码。

Sneakers manufactured in China are sold to consumers across the globe via platforms including Amazon and AliExpress. Consumer foot shapes vary across different countries, resulting in distinct requirements for properly fitting sneaker sizes. By integrating a 3D foot scanning and measurement algorithm model into online shopping malls to collect foot shape data of Thai consumers, including foot length, foot width, arch height, and metatarsal circumference, we conduct 3D model reconstruction and foot-shoe data matching to analyze and predict the optimal sneaker size tailored to Thai consumers’ foot dimensions. This solution helps enterprises engaged in sneaker export businesses reduce the return and exchange rate caused by ill-fitting footwear, as well as lower freight costs incurred during the return and exchange process.
1. Data Collection: Before purchasing sneakers, Thai consumers take three photos of their feet using their mobile phones and upload them to the system.
2. Data Processing: For the photos uploaded by Thai consumers, the 3D foot scanning and measurement algorithm model is invoked. First, the YOLO image recognition model is used to identify the consumer’s foot shape images, followed by the RCF edge detection algorithm to fully extract the complete foot shape from the photos. Based on the size data filled in by the user, 3D modeling is performed to restore the complete morphology of the foot. If image gaps arise during the reconstruction process, adjustments based on a size deviation of ±0.5 are applied, with a threshold limit of ±1, to restore the 3D foot model with the smallest gap.
3. Data Analysis and Prediction: Based on the results of the 3D foot modeling, the actual foot dimension data of the consumer is calculated through proportional analysis using dichotomy rules. By referring to the international standard foot-shoe size correspondence table, the most appropriately fitting sneaker size corresponding to the consumer’s foot dimensions is predicted.

搜集汇总

背景与挑战

背景概述

该数据集包含泰国地区消费者的脚型数据，用于预测运动鞋尺码，涵盖足长、足宽、足弓高和跖围等多维度测量指标，共752条记录并每年更新。通过三维足部扫描和算法模型，数据集支持运动鞋出口企业优化尺码匹配，降低退换货率和成本，提升销售效率。