Polarization Camera HDR Reconstruction via Channel Selection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current imaging devices have limited dynamic range, leading to loss of detail in bright or dark areas of images, and existing methods to increase dynamic range, such as capturing multiple exposures, result in motion blur, depth of field issues, and sensor noise, while using polarization images can be corrupted by over/underexposed pixels.
Innovation Solution
A method for reconstructing image irradiance using a polarization camera that captures images at different angles, selects the color channel with the least number of over/underexposed pixels, estimates degree and angle of linear polarization based on prior probability distributions, and combines polarized irradiance values to generate a reconstructed image with expanded dynamic range.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If multiple exposures are used to increase dynamic range, then dynamic range is improved, but motion blur and depth of field issues occur
Solution Approach 1:
The patent segments the polarization images into multiple color channels (e.g., red, green, blue) and processes each channel separately. By dividing the data into distinct channels, the system can select the optimal channel with the least over/underexposure and perform HDR reconstruction on that channel, avoiding the need for multiple exposures while maintaining image quality.
Solution Approach 2:
The patent changes the parameter of polarization angle by capturing images at multiple polarization angles (e.g., 0°, 45°, 90°, 135°) instead of using multiple exposures. This parameter change allows the system to achieve HDR effect through polarization modulation rather than exposure time variation, eliminating motion blur and depth of field issues.
2Illumination intensity
If polarization images are used to expand dynamic range, then dynamic range is improved, but over/underexposed pixels corrupt the image
Solution Approach 1:
The patent applies local quality by evaluating each color channel's over/underexposure characteristics individually and selecting the channel with the best quality for HDR reconstruction. Instead of treating all pixels uniformly, the system identifies local regions (color channels) with different exposure qualities and processes them accordingly, ensuring high measurement precision in the selected channel.
Solution Approach 2:
The patent creates a synthesized HDR image by copying and combining information from multiple polarization images. The system generates a virtual HDR representation by integrating data from different polarization angles and color channels, effectively copying the scene's radiance information across multiple measurements to reconstruct accurate pixel values.
3Reliability
If single exposure is used with polarization images, then motion blur is avoided, but dynamic range is limited
Solution Approach 1:
The patent introduces another dimension by utilizing polarization angle as an additional parameter for capturing scene information. Instead of relying solely on exposure time (temporal dimension), the system captures multiple polarization angles simultaneously, adding a new dimension to the data space. This enables HDR reconstruction from a single exposure by exploiting the polarization dimension rather than the temporal dimension.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The method produces a reconstructed image with a higher dynamic range than the imaging device's capabilities, minimizing the impact of over/underexposed pixels and providing more accurate image classification for computer vision applications.
Implementation Method 1
The image of the scene includes a plurality of polarization images corresponding to different angles of polarization
Implementation Method 2
obtaining an inverse camera response function for the one color channel for mapping pixel values of the pixels to corresponding polarized irradiance values
Implementation Method 3
the combining of the polarized irradiance values includes applying a weight to the polarized irradiance values
Implementation Method 4
estimating a degree of linear polarization (DOLP) and angle of linear polarization (AOLP) for the pixel based on a prior probability distribution of the DOLP and AOLP
Data Source
AI summary
Systems and method for expanding a dynamic range associated with an image are disclosed. The method includes capturing an image of a scene via an imaging device using a single exposure. The image of the scene includes a plurality of polarization images corresponding to different angles of polarization, and each of the polarization images comprise a plurality of color channels. The method further includes determining a criterion for each of the plurality of color channels; selecting one color channel of the plurality of color channels based on determining of the criterion; generating a reconstructed image irradiance for the one color channel based on pixels in two or more of the plurality of polarization images obtained for the one color channel; and outputting a reconstructed image with the reconstructed image irradiance.