Video Surveillance Object Recognition Calibration Function
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video surveillance systems face challenges in accurately recognizing objects due to variations in images captured under different scenarios, such as different lighting conditions and camera locations, which affect the similarity analysis and decrease recognition accuracy.
Innovation Solution
A system that includes a camera and a processor to capture and process images, using a calibration function to adjust similarity degrees based on correlations between images captured under different scenarios, allowing for improved object recognition by comparing images against a threshold for authentication.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If images are captured under different scenarios (different lighting conditions, camera locations), then the system can monitor more environmental conditions, but the similarity analysis accuracy decreases
Solution Approach 1:
The patent applies parameter changes by introducing a calibration function that adjusts similarity assessment parameters based on scenario-specific characteristics. The system captures images under different scenarios (varying lighting, camera locations, angles) and uses calibration functions to normalize the similarity comparison, allowing the system to adapt to diverse environmental conditions while maintaining recognition accuracy.
2Productivity
If a fixed similarity threshold is used for object recognition, then the recognition process is simple and fast, but the recognition accuracy decreases under varying capture conditions
Solution Approach 1:
The patent implements dynamics by transforming the fixed similarity threshold into a dynamic calibration function. Instead of using a static threshold value, the system applies scenario-specific calibration functions that adjust the similarity assessment adaptively. This allows the recognition process to remain efficient while improving accuracy under varying capture conditions through dynamic parameter adjustment.
3Ease of operation
If images of the same object captured under different scenarios are compared directly, then the comparison process is straightforward, but the features differ and influence similarity determination negatively
Solution Approach 1:
The patent introduces an intermediary calibration function between the direct comparison of images captured under different scenarios. This calibration function acts as a mediator that normalizes the features extracted from images taken under varying conditions (different lighting, camera locations, angles) before performing similarity determination, thereby maintaining comparison simplicity while improving accuracy.
Data Source
AI summary
A system for object recognition are provided in the present disclosure. The system may obtain a first image of an object that is captured by a camera configured to capture one or more images for use in an object recognition process under a first scenario; obtain a second image of the object that is captured under a second scenario; assess a degree of similarity between the first image of the object and at least one sample image; and determine a calibration function to calibrate the degree of similarity between the first image of the object and the at least one sample image based at least on a correlation between the second image of the object and the at least one sample image, wherein the calibration function is to be applied in association with the one or more images captured by the camera in the object recognition process.