Surveillance Search Apparatus Using Attribute Weighting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing surveillance camera systems face inefficiencies in searching large numbers of images for a specific person, especially when the distribution of the person's attributes is not uniform, leading to slow search times.
Innovation Solution
A search apparatus that determines multiple attributes of a designated object and those different from it, using a combination of attribute extraction, identification performance evaluation, and weighting factors to efficiently search for the object across multiple surveillance camera images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If simple similarity summation of multiple attributes is used to determine collective similarity, then the search method is simple and easy to implement, but search efficiency deteriorates when attribute distribution is not uniform
Solution Approach 1:
The patent changes the parameter of similarity determination from simple summation to weighted summation, where weights are assigned based on the distribution characteristics of each attribute. This allows the system to adapt to non-uniform attribute distributions by giving higher importance to attributes with more discriminative power, thereby improving search efficiency while maintaining implementation feasibility
Solution Approach 2:
The patent introduces dynamic weighting factors that are determined based on the actual distribution of attributes in the database. Instead of using fixed equal weights, the system dynamically adjusts the weight of each attribute according to its discriminative capability, making the search method adaptive to different data scenarios and improving overall search efficiency
2Device complexity
If equal weights are assigned to all attributes, then the search algorithm is simple, but search accuracy deteriorates when some attributes are more distinguishing than others
Solution Approach 1:
The patent transforms the parameter of attribute weighting from uniform to non-uniform by introducing weight coefficients that reflect the discriminative power of each attribute. This change allows the system to prioritize attributes that are more effective for identification, thereby improving search accuracy without significantly increasing algorithmic complexity
Solution Approach 2:
The patent applies different weights to different attributes based on their local characteristics and discriminative power. Instead of treating all attributes equally, the system assigns higher weights to attributes with better identification performance and lower weights to less discriminative attributes, achieving improved accuracy through localized quality adjustment
3Measurement precision
If more attributes are evaluated with higher precision, then search accuracy is improved, but computational complexity and processing time increase
Solution Approach 1:
The patent changes the parameter of attribute evaluation by introducing weighted scoring that prioritizes attributes with higher discriminative power. This allows the system to achieve better search accuracy with fewer attributes by focusing computational resources on the most informative attributes, thereby reducing overall processing time
Solution Approach 2:
The patent extracts and prioritizes the most discriminative attributes by evaluating their distribution characteristics and assigning higher weights to them. This extraction of key informative attributes allows the system to achieve high search accuracy without processing all attributes with equal precision, thus reducing computational burden and processing time
Data Source
AI summary
A search apparatus comprises a determining unit configured to determine a plurality of attributes of an object designated in a first image, and determine a plurality of attributes of an object that is different from the designated object; and a search unit configured to search a second image for the designated object in accordance with a search condition based on the plurality of attributes of the designated object and the plurality of attributes of the object that is different from the designated object.