Customer Flow Statistical Method Using Facial Recognition
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for customer flow statistics are inaccurate due to the inclusion of non-relevant personnel, such as staff, which interferes with the data and reduces the accuracy of customer flow calculations.
Innovation Solution
A statistical method and apparatus that identifies facial areas in video data and matches them with pre-set facial information to differentiate between relevant and irrelevant individuals, using a counter to determine the number of unsuccessfully matched facial areas as the customer flow, thereby isolating the customer flow from unrelated persons.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional customer flow statistics method is used to count all persons entering a store, then the counting process is simple, but the accuracy of customer flow statistics is reduced due to inclusion of non-customer personnel such as staff
Solution Approach 1:
The patent segments the population into two distinct groups: customers and non-customers (staff). By using facial recognition technology to identify and separate these groups, the system can selectively count only customers, thereby improving measurement precision without requiring a complete overhaul of the counting mechanism
Solution Approach 2:
The patent introduces facial recognition technology as an intermediary mechanism between the person detection system and the counting system. This intermediary layer filters out non-customer personnel by comparing facial features against a database of known staff members, allowing the system to accurately distinguish and count only customers
2Measurement precision
If facial recognition technology is implemented to differentiate customers from staff, then the accuracy of customer flow statistics is improved, but the device complexity and processing time increase
Solution Approach 1:
The patent performs preliminary action by pre-storing facial recognition data of all staff members in a database before the actual customer flow counting begins. This preparation work is done in advance, so that during the counting process, the system only needs to perform rapid comparison operations rather than building the recognition database from scratch, significantly reducing processing time
Solution Approach 2:
The patent applies local quality by optimizing the facial matching process to work efficiently in the specific context of customer flow monitoring. The system tailors the recognition algorithm to work with the specific characteristics of entrance/exit video data, using localized processing strategies that reduce overall processing time while maintaining high accuracy
Data Source
AI summary
Disclosed is a statistical method, apparatus and device for customer flow. The method includes: acquiring video data on which the statistics is to be made (S101), identifying facial areas in the video data (S102), matching the identified facial areas with pre-set facial information, wherein the pre-set facial information may be facial information of a person unrelated to the customer flow (for example, a staff, etc.); determining the number of successfully matched facial areas, to obtain the customer flow without unrelated persons (S103). It can be seen that the method removes the interference from unrelated persons and improves the accuracy of customer flow statistics.