Image Processor Calibration Map for Depth Map Accuracy
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Image processors face challenges in evenly reflecting the blurring phenomenon in the background part of images and in properly calibrating depth maps according to illuminance, especially when using CMOS image sensors for generating depth maps.
Innovation Solution
An image processor is designed with a calibration map generator to create a calibration map representing pixel value characteristics, and a calibrator to adjust disparity values for each pixel area based on the calibration map, ensuring even blurring reflection and illuminance adaptation, using a system that includes an image sensor and an image processor to generate and calibrate depth maps.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the image processor uses a Bokeh application to generate depth map, then the depth map can be generated for background blur effect, but the blurring phenomenon is not evenly reflected in the background part of the image
Solution Approach 1:
The patent applies parameter changes by adjusting the disparity value through calibration based on pixel area characteristics. The calibrator modifies the disparity value using calibration data that accounts for variations across different pixel areas, ensuring uniform blurring effect while maintaining depth map accuracy.
Solution Approach 2:
The patent implements local quality by applying different calibration adjustments to different pixel areas. The calibration process considers the specific characteristics of each pixel area (such as position in the image sensor) and applies localized corrections to achieve uniform blurring across the entire image background.
2Productivity
If the image processor generates depth map without calibration, then the processing speed is faster, but the depth map cannot be properly calibrated according to illuminance
Solution Approach 1:
The patent applies preliminary action by pre-generating calibration data that captures the relationship between pixel area characteristics and disparity values under different illuminance conditions. This calibration data is prepared in advance and stored for quick reference during actual depth map generation, enabling fast processing without sacrificing calibration accuracy.
Solution Approach 2:
The patent introduces an intermediary calibration map that mediates between the raw disparity values and the final depth map. This calibration map contains pre-computed correction factors that adjust disparity values based on pixel area and illuminance conditions, allowing the system to maintain both speed and accuracy.
3Measurement precision
If the calibration map is generated for each illuminance condition, then the depth map calibration accuracy is improved, but the device complexity increases
Solution Approach 1:
The patent applies universality by creating a unified calibration framework that handles multiple illuminance conditions through a single calibration map structure. The calibration map is designed to be applicable across different lighting conditions, reducing the need for separate calibration data for each illuminance level and simplifying the overall system.
Solution Approach 2:
The patent uses parameter changes by incorporating illuminance level as a variable in the calibration process. Instead of creating separate calibration maps for each illuminance condition, the system adjusts the calibration parameters based on the current illuminance level, allowing a single calibration framework to adapt to different lighting conditions.
Data Source
AI summary
Disclosed is an image processor and an image processing system including the image processor, and the image processing system may include an image sensor including a plurality of pixel areas, and configured to generate pixel values from the plurality of pixel areas, and an image processor configured to, based on the pixel values and a calibration map, calibrate a disparity value for each pixel area, and generate a depth map.