M×N-Cell Image Sensor HDR+ Noise and Ghosting Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current HDR technologies face challenges such as time delays between exposure images, leading to ghosting or blurring, especially when the camera is shaken or objects are moving. Additionally, the fused images often contain significant noise due to distinct exposure levels and noise characteristics in each image.
Innovation Solution
The implementation of HDR+ using M×N-cell image sensors, which capture a plurality of frames using a short-exposure setting. These sensors merge the frames by identifying pixel groups and combining M×N pixels into super-pixels, followed by tone mapping to reduce the dynamic range of the HDR image into a low dynamic range (LDR) image.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Illumination intensity
If frame-based HDR algorithm is used to capture multiple exposure images, then dynamic range is improved, but time delay between images causes ghosting or blurring
Solution Approach 1:
The image sensor is divided into multiple pixel groups (M×N cells), where each pixel group contains multiple pixels that capture the same color information simultaneously. This segmentation allows parallel processing of multiple exposure levels within a single frame, eliminating the time delay between exposures while maintaining high dynamic range.
Solution Approach 2:
Multiple pixel groups with different exposure levels are merged within a single frame capture. The M×N pixels in each pixel group are combined to form super-pixels in the HDR image, allowing simultaneous capture of multiple exposure levels without temporal separation, thus eliminating ghosting and blurring caused by time delay.
2Illumination intensity
If multiple frames with distinct exposure levels are merged, then dynamic range is improved, but noise in fused image increases
Solution Approach 1:
Each pixel group is configured with pixels of the same color that capture the same color information, allowing local optimization of noise reduction. By processing each pixel group independently and then merging them, the system can apply noise reduction techniques tailored to each local region's exposure level and noise characteristics.
Solution Approach 2:
Multiple pixel groups capture redundant color information at different exposure levels. This redundancy allows the system to select the best quality data from multiple copies and discard noisy or degraded information, thereby reducing overall noise in the fused HDR image while maintaining dynamic range.
3Illumination intensity
If M×N pixels in each pixel group are merged into super-pixels, then dynamic range is improved, but image resolution is reduced
Solution Approach 1:
The system transitions from capturing data in spatial dimensions to processing data in a combined spatial-temporal dimension. By merging M×N pixels into super-pixels while maintaining multiple exposure levels, the system achieves high dynamic range in the vertical dimension (exposure levels) while preserving horizontal spatial resolution through the M×N pixel grouping structure.
Data Source
AI summary
This application describes method and apparatus for HDR+ using M×N-cell sensors. An example apparatus includes a plurality of pixel groups, each pixel group having M×N pixels configured with color filters of a same color such that the M×N pixels in each pixel group capture a same color. M and N are integers greater than one, and each pixel is represented by a first number of bits. The example apparatus further include one or more processors configured to: capture, through the plurality of pixel groups, a plurality of frames of a scene using a short-exposure setting; merge the plurality of frames into an HDR image; and perform tone mapping on the HDR image to reduce a dynamic range of the super-pixels of the HDR image into a low dynamic range (LDR) image, wherein each pixel in the LDR image is represented with the first number of bits.