HDR Image Processing Pixel Validity Map Bandwidth Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image processing systems consume significant power and bandwidth when generating high dynamic range (HDR) images, particularly due to the need to write and read over-exposed pixels from long exposure images and under-exposed pixels from short exposure images, which do not contribute to the final HDR image.
Innovation Solution
The system suppresses or replaces over-exposed pixels from long exposure images and under-exposed pixels from short exposure images based on threshold values, using pixel validity maps to indicate which pixels are suppressed, allowing only valid pixels to be stored and processed, and reconstructing suppressed pixels from corresponding pixels in other exposure images during output.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all pixels from multiple exposure images are stored and processed to generate HDR images, then image quality is maintained, but power consumption and bandwidth usage increase significantly
Solution Approach 1:
The patent extracts and identifies only the valid pixels from each exposure image that actually contribute to the final HDR image. By separating valid pixels from invalid ones using pixel validity maps, the system processes and stores only the necessary data, eliminating waste while preserving image quality.
Solution Approach 2:
The patent changes the parameter of pixel data representation by introducing pixel validity maps that indicate which pixels are valid or invalid. This parameter change allows the system to selectively process only valid pixels, reducing power consumption without compromising the quality of the final HDR image.
2Loss of information
If all pixels from multiple exposure images are stored in memory, then complete image data is preserved, but bandwidth consumption and memory usage increase
Solution Approach 1:
The patent extracts only the valid pixel data from each exposure image for storage in memory, using pixel validity maps to identify which pixels should be retained. This extraction process preserves all necessary image information while eliminating redundant data, thereby reducing bandwidth consumption during data transfer.
Solution Approach 2:
The patent discards invalid pixels that do not contribute to the final HDR image during the storage phase. These invalid pixels are effectively removed from the data stream, reducing memory usage and bandwidth consumption, while the system recovers the complete image quality by selectively combining only valid pixels from multiple exposures during HDR generation.
3Loss of information
If over-exposed and under-exposed pixels are processed, then all possible image information is captured, but processing time and computational resources are wasted
Solution Approach 1:
The patent extracts and processes only valid pixels from each exposure image by using pixel validity maps to identify which pixels fall within acceptable exposure ranges. This selective processing captures all necessary image information while eliminating waste of computational resources on invalid pixels.
Solution Approach 2:
The patent applies partial action by processing only the necessary subset of pixels (valid pixels) rather than all pixels from each exposure image. This approach captures sufficient image information for high-quality HDR generation without the excessive processing time and computational resources that would be required to process every pixel.
Data Source
AI summary
Techniques are described herein for processing image data. For instance, a technique can include obtaining a first image having a first exposure time and a second image having a second exposure time that is greater than the first exposure time. The technique can further include determining at least one of: that a first pixel of the first image has a first pixel value below a first threshold value; or that a second pixel of the second image has a second pixel value above a second threshold value; suppressing at least one of the first pixel or the second pixel based on the determination so as to prevent storing the first pixel or the second pixel; replacing the suppressed first pixel or second pixel based on a another pixel value; and outputting the first image or the second image, the first image or the second image including the another pixel.