Image Processing Apparatus Optical Distortion Recognition Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image recognition systems experience decreased recognition accuracy when dealing with images captured using optical systems that introduce significant distortion, as they are not trained on distorted images, leading to increased processing time and resource usage for distortion correction, and large classifier networks for each optical system.
Innovation Solution
An image processing apparatus that acquires information about an optical system and estimates recognition accuracy in peripheral areas, allowing the system to determine whether to use target pixels or areas for recognition based on pre-defined conditions, thereby reducing the need for distortion correction and optimizing processing time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If distortion correction is performed for each captured image, then recognition accuracy is improved, but processing time increases and calculation resources are consumed
Solution Approach 1:
The system performs preliminary actions by acquiring optical system information and calculating distortion characteristics in advance, storing them for later use. This allows the recognition apparatus to directly use pre-calculated distortion data without performing time-consuming distortion correction for each image, thereby maintaining recognition accuracy while reducing processing time.
2Measurement precision
If distortion correction is performed for each captured image, then recognition accuracy is improved, but calculation resource usage increases
Solution Approach 1:
The image processing apparatus performs preliminary calculations of distortion characteristics and stores them in advance. When recognition is needed, the system directly retrieves and uses these pre-calculated values instead of performing intensive real-time distortion correction calculations, significantly reducing calculation resource consumption while maintaining recognition accuracy.
3Measurement precision
If learning is performed for each optical system, then recognition accuracy for that optical system is improved, but the network of classifiers becomes large
Solution Approach 1:
Instead of creating separate classifier networks for each optical system, the system changes the approach by using optical system information as input parameters. The recognition apparatus acquires information about the optical system and uses this to select or adjust recognition parameters, allowing a single network to handle multiple optical systems without becoming excessively large.
Solution Approach 2:
The system implements universality by creating a single recognition apparatus that can handle multiple optical systems. By acquiring optical system information and using it to guide recognition parameter selection, the same network structure serves multiple optical systems, avoiding the need for separate specialized networks for each system.
Data Source
AI summary
An image processing apparatus includes a memory storing instructions, and a processor configured to execute the instructions to acquire information about an optical system included in an image pickup apparatus configured to capture an image, acquire a change amount in the information about the optical system in a peripheral area of a target pixel in the image based on the information about the optical system, and acquire, based on the change amount, information about recognition accuracy in the peripheral area by a recognition apparatus configured to perform image recognition.