A customer dress attribute analysis method and system and a computer device

Abstract

The invention discloses a customer dress attribute analysis method and system and a computer device. The method comprises the following steps: obtaining pictures or / and video data in a store and extracting a single picture from the pictures; using the trained first-depth convolution neural network model to perform the customer humanoid detection on the single picture, and obtaining the customer humanoid frame position; Carrying out picture interception according to the position of the humanoid frame to obtain the whole humanoid picture of the customer; using The trained second-depth convolution neural network model to analyze the customer dressing attributes of the whole customer humanoid picture, and obtaining the confidence of the customer dressing attributes. According to the attributeselection strategy, determining the customer dress attributes, outputting the corresponding attributes and confidence. The invention can quickly and accurately detect the dress attributes of the customers in the store, and then statistics and analyzes the dress styles, dress collocation and other information of the customers in the store, which is helpful for the store to formulate the sales strategy and promote the sales volume.

Description

technical field [0001] The present invention relates to the technical field of big data processing, in particular to a method, system and computer equipment for analyzing customer attributes. Background technique [0002] At present, although online stores have a certain impact on offline stores, offline stores such as various large shopping malls, commodity stores, supermarkets, etc. still play an important role in people's daily life. At present, a prominent problem in various offline stores is that they do not perceive customers. On the one hand, there are no professionals or labor costs to analyze customer attributes. On the other hand, the cameras installed in offline stores are mainly used for video surveillance to prevent theft. It happened and provided strong video evidence when necessary, but did not conduct corresponding data mining through in-store video data, and mine valuable information from a large number of customers who came to the store. Contents of the i...

AI Technical Summary

Problems solved by technology

At present, a prominent problem in various offline stores is that they do not perceive customers. On the one hand, there are no professionals or labor costs to analyze customer attributes. On the other hand, the cameras installed in offline stores are mainly used for video surveillance to prevent theft. occurred and provide strong video evidence when necessary, but did not conduct corresponding data mining through in-store video data, and mine valuable information from a large number of customers who came to the store

Images