Image Capture Control Apparatus Moving Body Exposure Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image capture apparatuses face issues with degraded image quality due to rapid changes in exposure, such as when a high-luminance moving object enters the frame, leading to insufficient exposure for other areas, and changes in object distance or white balance.
Innovation Solution
An image capture control apparatus that includes a moving body detection unit, prediction unit, statistical information generation unit, and control information generation unit to predict and adjust exposure and focus settings based on statistical analysis of pixel values, excluding the predicted moving body area from the processing area to maintain optimal exposure and focus for the entire frame.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If exposure control is performed based on the entire frame including moving bodies, then the moving body is properly exposed, but other areas in the frame become underexposed
Solution Approach 1:
The frame is segmented into multiple regions: a first region containing the moving body and a second region excluding the moving body. Exposure control is performed separately for each region, allowing the moving body to be properly exposed while maintaining appropriate exposure levels in other areas. This segmentation enables independent exposure adjustment for different spatial zones within the same frame.
Solution Approach 2:
Different exposure qualities are applied to different regions of the frame based on local requirements. The first region (with moving body) receives one exposure setting optimized for capturing the moving subject, while the second region (other areas) receives a different exposure setting optimized for static background elements. This local quality approach ensures each region achieves its optimal exposure independently.
2Manufacturing precision
If mechanical components such as camera platform are used for panning and zooming, then image quality is maintained, but cost and power consumption increase
Solution Approach 1:
The patent replaces mechanical panning and zooming systems with electronic image processing techniques. Instead of physically moving the camera platform or lens to achieve panning and zooming effects, the system uses digital image processing to crop, scale, and reposition image data electronically. This substitution eliminates the need for mechanical components while maintaining the functional outcomes, thereby reducing cost and power consumption.
3Use of energy by stationary object
If electronic trimming is used for panning and zooming, then cost and power consumption are reduced, but image quality is degraded
Solution Approach 1:
The image processing is segmented into distinct stages: capturing a wider frame, identifying the moving body region, and then electronically trimming/cropping to the relevant area. This segmentation allows the system to capture sufficient detail in the initial wide frame while enabling high-quality electronic zooming and panning in subsequent processing stages, thereby maintaining image quality despite using electronic rather than mechanical methods.
Data Source
AI summary
To improve image quality of image data.A moving body detection unit detects, as a moving body area, an area of a moving body from an input frame every time the input frame captured by an image capture unit is input. A prediction unit predicts, as a prediction area, an area corresponding to the moving body area in a next input frame following the input frame from the moving body area every time the input frame is input. A statistical information generation unit that generates statistical information by performing predetermined statistical processing on each of pixel values of pixels in either an area excluding the predicted prediction area from a specific processing area in the input frame or an area that overlaps the prediction area in the processing area every time the input frame is input. A control information generation unit that generates predetermined control information for controlling the image capture unit based on the statistical information every time the statistical information is generated. A controller that controls the image capture unit and causes the image capture unit to capture the next input frame based on the control information every time the control information is generated.