Street Level Image Face Detection for Advertisement Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing street level image analysis systems fail to accurately identify and differentiate between human faces present in advertisements and those of individuals in real-life scenarios, leading to potential privacy concerns and inefficiencies in managing advertisements across multiple locations.
Innovation Solution
A method and system that utilize a processor to detect and compare faces in street level images, determining if they are similar and likely part of an advertisement by analyzing image features and locations, allowing for the identification and storage of advertisement boundaries for potential replacement.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If face detection technology is applied to street level images to protect individual privacy, then privacy protection is improved, but advertisement identification capability deteriorates
Solution Approach 1:
The system applies different processing rules to different regions within detected face areas. By analyzing spatial relationships between multiple detected faces and considering geographic location data, the system determines whether faces belong to advertisements or individuals, applying appropriate handling (no blurring for ads, blurring for individuals) to each specific case.
Solution Approach 2:
Instead of automatically blurring all detected faces to protect privacy, the system inverts the approach by first assuming potential advertisement content and then verifying through comparison analysis. If faces are determined to be from advertisements based on similarity across multiple locations, no blurring is applied, thus preserving advertisement identification capability while maintaining privacy protection.
2Reliability
If all detected faces are blurred to protect privacy, then privacy protection is improved, but advertisement management efficiency deteriorates
Solution Approach 1:
The system implements selective face handling by analyzing the relationship between multiple detected faces and their geographic locations. Faces identified as part of advertisements across multiple locations are preserved without blurring, while faces of actual individuals are blurred, achieving both privacy protection and advertisement management goals.
Solution Approach 2:
The system performs preliminary comparison analysis of detected faces across multiple images before finalizing privacy protection measures. By pre-identifying advertisement faces through similarity comparison and location analysis, the system can apply appropriate processing (blurring or preservation) efficiently without requiring manual review of each face.
3Measurement precision
If face comparison across multiple images is performed to identify advertisements, then advertisement identification accuracy is improved, but processing complexity increases
Solution Approach 1:
The system segments the face identification process into distinct stages: initial face detection in each image, extraction of facial features, comparison of features across multiple images, geographic location-based relationship analysis, and final determination of advertisement status. This segmentation reduces processing complexity by handling each aspect separately with optimized algorithms.
Solution Approach 2:
The system introduces geographic location data as an intermediary element to facilitate advertisement identification. By comparing not only facial features but also the spatial relationships between detected faces across different locations, the system can more accurately distinguish advertisement faces from individual faces without requiring overly complex direct image comparison algorithms.
Data Source
AI summary
A system and method is provided wherein, in one aspect, a processor determines whether multiple street level images have captured a nearly-identical face. If so, the images are processed to determine whether the face appears to be part of an advertisement. Once it is determined that the face is displayed on an advertisement, the boundaries of the advertisement may be determined and the location of the advertisement is stored for future use, e.g., potentially replacing the advertisement in the image with a different advertisement.