Image Stitching Calibration via Contrast Reference and Distribution Transform
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional camera systems struggle to produce high-quality stitched images due to differences in automatic exposure adjustment, luminance, and chroma between images captured by adjacent image sensors, resulting in noticeable junctions and low image quality.
Innovation Solution
An image calibrating method that calculates contrast values, sets the image with maximal contrast as a reference, and adjusts target images using a transform function derived from probability distribution information to match the luminance and chroma of the reference image, ensuring uniformity and reducing chromatic aberration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If multiple image sensors are disposed adjacently to capture images in different directions, then the image monitoring range is expanded, but the luminance and chroma differences between images cause obvious junctions and reduce stitching quality
Solution Approach 1:
The patent applies parameter changes by adjusting the luminance and chroma parameters of target images through transform functions. The system calculates probability distribution information of the reference image and target images, then obtains transform functions to adjust pixel values, thereby changing the luminance and chroma parameters to match between different image sensors and eliminate obvious junctions in stitched images
Solution Approach 2:
The patent uses an intermediary approach by introducing a reference image as a mediator. The reference image serves as a standard against which target images are compared and adjusted. The transform functions act as intermediaries to transfer the luminance and chroma characteristics from the reference image to the target images, enabling seamless stitching without direct manipulation of each sensor's output
2Reliability
If automatic exposure adjustment is applied by different image sensors, then each sensor optimizes its captured image, but the exposure differences create noticeable junctions in the stitching image
Solution Approach 1:
The patent changes the exposure-related parameters (luminance and chroma) of target images using transform functions derived from probability distribution information. This adjusts the exposure characteristics of target images to match the reference image, eliminating noticeable junctions while preserving the optimization benefits of automatic exposure adjustment for each sensor
Data Source
AI summary
An image calibrating method for stitching images includes calculating contrast values of a plurality of images, and comparing the contrast values to set the image with maximal contrast value as a reference image; others of the plurality of images are set as target images. The image calibrating method further includes calculating cumulative distribution functions of the reference image and the target images respectively, and obtaining a transform function according to the cumulative distribution functions of the reference image and the target images for pixel adjustment of the target images. Further, a camera with an image calibrating function and a related image processing system are disclosed in the present invention.