Image Sensor Tone Mapping via Luminance Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image sensing devices face limitations in generating high dynamic range (HDR) images due to the limitations of global and local tone mapping methods, which either fail to improve contrast effectively or require excessive hardware resources.
Innovation Solution
An image sensing device employing a semi-global tone mapping method that combines global and local tone mapping techniques, using a tone mapping module to generate tone mapping information based on histograms divided by luminance, and an HDR image generation module to produce HDR images by leveraging both methods.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If global tone mapping method is used, then processing simplicity is maintained, but local contrast improvement is insufficient
Solution Approach 1:
The patent divides the histogram into multiple regions based on luminance thresholds, creating separate tone mapping curves for different luminance ranges. This segmentation allows the system to apply different processing strategies to different parts of the image, improving local contrast while maintaining overall processing efficiency.
Solution Approach 2:
The patent generates multiple tone mapping curves corresponding to different luminance regions (e.g., dark, medium, bright regions). Each curve is optimized for its specific luminance range, enabling local contrast improvement in different parts of the image while maintaining global tone mapping efficiency.
2Manufacturing precision
If local tone mapping method is used, then local contrast improvement is enhanced, but hardware resources increase excessively
Solution Approach 1:
The patent applies local tone mapping only to specific luminance regions that require enhancement, rather than processing the entire image with full local tone mapping complexity. By focusing computational resources on partial regions (e.g., only dark or only bright regions), the system achieves local contrast improvement while reducing overall hardware resource consumption.
Solution Approach 2:
The patent changes the parameter of tone mapping curve selection based on luminance thresholds. Instead of using a fixed complex local tone mapping algorithm for all pixels, the system selects from multiple pre-generated curves based on pixel luminance values, reducing computational complexity while maintaining local contrast improvement.
3Productivity
If single tone mapping curve is used, then processing efficiency is maintained, but contrast improvement across different luminance regions is limited
Solution Approach 1:
The patent segments the luminance range into multiple regions and generates a separate tone mapping curve for each region. This allows the system to process different luminance ranges with optimized curves, improving contrast across all regions while maintaining efficiency through the use of multiple simple curves rather than one complex algorithm.
Solution Approach 2:
The patent creates a universal set of tone mapping curves that can be applied to different luminance regions. Each curve serves multiple purposes by being applied to specific luminance ranges across the entire image, allowing efficient processing while achieving comprehensive contrast improvement.
Data Source
AI summary
Disclosed are an image sensing device and an operating method thereof, and the image sensing device may include: an image sensor including a plurality of pixels and suitable for generating an image based on incident light; and an image processor suitable for generating a high dynamic range (HDR) image based on the image and two or more pieces of tone mapping information, which are divided according to luminance.