Stereo Camera State Diagnosis Using Disparity Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing video monitoring systems face reliability issues due to camera abnormalities, such as dirt or intentional tampering, which can lead to inaccurate state classification and reduced detection accuracy, especially in challenging environments like dark places or uniform luminance conditions, compromising security.
Innovation Solution
A photographing device equipped with a stereo camera, image processing unit, diagnostic pattern database, state diagnosis unit, and camera control unit that acquires disparity information, stores diagnostic patterns, and performs state diagnosis to determine camera states and adjust camera settings, including illumination and posture control.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If disparity image data is used to determine camera normal/abnormal states, then classification works in normal lighting conditions, but it becomes difficult to classify camera states in dark places or uniform luminance conditions where disparity image data cannot be obtained
Solution Approach 1:
The system performs preliminary classification of camera images into multiple types (first type with normal disparity, second type without disparity, third type with abnormal disparity) before final state determination. This preliminary action allows the system to handle different lighting conditions appropriately by selecting suitable classification methods for each image type, thereby resolving the contradiction between measurement precision and adaptability.
Solution Approach 2:
The system changes the parameter used for state determination based on image characteristics. For images where disparity information is unavailable or abnormal, the system switches from using disparity-based parameters to using alternative parameters such as image brightness, contrast, or other visual features, enabling state determination across various lighting conditions while maintaining accuracy.
2Reliability
If the camera is classified as normal based on disparity information, then the system operates normally, but dirt on the lens may be neglected and not determined as abnormal
Solution Approach 1:
The system performs preliminary classification to identify images with abnormal disparity patterns that may indicate lens contamination. By categorizing images into different types before final determination, the system can flag potential dirt conditions for further analysis, preventing false normal classifications and improving reliability.
Solution Approach 2:
The system uses feedback from disparity analysis to trigger additional verification steps. When disparity information suggests potential abnormalities (such as unexpected patterns that could indicate lens dirt), the system feeds this information back into the determination process to re-evaluate the camera state, ensuring dirt is not neglected and improving detection reliability.
3Measurement precision
If a stereo camera is used to measure distance for stable recognition, then recognition accuracy improves, but the system becomes vulnerable to intentional tampering such as changing photographing angle or placing blocking objects
Solution Approach 1:
The system performs preliminary classification of captured images to detect signs of tampering such as abnormal disparity patterns, unexpected blocking objects, or inconsistent depth information. By identifying these harmful factors before final recognition, the system can reject compromised images and maintain recognition accuracy while being resistant to intentional tampering attempts.
Data Source
AI summary
A photographing device for photographing a monitored area has a state diagnosis of a camera using a stereo camera configured by two or more cameras. A diagnostic pattern database is provided in which information on a plurality of images having different photographing conditions obtained by the stereo camera are stored in advance as diagnostic patterns. Disparity information is acquired based on image data from the stereo camera and a state diagnosis of the camera is performed by determining whether the camera is abnormal or normal based on the disparity information with reference to the information on the plurality of images having different photographing conditions stored in the diagnostic pattern database. Upon receiving a result of the state diagnosis, at least one of illumination control, shutter control, and posture control of the camera or the stereo camera is performed.