CMOS Image Sensor Intensity Compensation Without Line Memories
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Digital cameras using CMOS image sensors (CIS) face challenges in maintaining image quality, especially in low luminance areas, as they often over-brighten bright regions and fail to preserve original colors, and existing techniques require significant hardware resources like line memories.
Innovation Solution
An apparatus and method that processes images using a CMOS image sensor, employing an image signal collection unit and an image correction unit to apply a compensation curve based on the mean intensity of the image, multiplying low-intensity areas with a high weight and high-intensity areas with a weight close to 1.0, thereby improving image quality without requiring line memories.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If Retinex technique is used to improve low luminance and image quality, then image quality in low luminance areas is improved, but hardware resources are significantly increased due to use of line memories
Solution Approach 1:
The patent extracts the essential function of Retinex technique (low luminance compensation) while removing the heavy hardware dependency (line memories). The image correction unit applies compensation curves to adjust intensity without requiring large memory resources, thus extracting the beneficial effect while eliminating the harmful hardware burden.
Solution Approach 2:
The patent creates a simplified copy of the Retinex processing approach that achieves similar low luminance compensation effects but with reduced hardware complexity. Instead of using actual line memories, the system uses an image correction unit with compensation curves that replicate the functional outcome with minimal hardware resources.
2Illumination intensity
If conventional techniques are used to improve luminance of images using CIS, then low luminance areas are brightened, but bright areas become over-brightened and original colors are not maintained
Solution Approach 1:
The patent applies different compensation strategies to different intensity regions. The image correction unit uses compensation curves that specifically target low luminance areas while preserving bright areas and color information. This local differentiation ensures that luminance enhancement is applied selectively where needed, maintaining color accuracy in both dark and bright regions.
3Adaptability or versatility
If Retinex technique with line memories is used, then wide dynamic range is achieved, but hardware cost and complexity increase significantly
Solution Approach 1:
The patent replaces the mechanical/memory-based Retinex implementation with a computational approach using an image correction unit. Instead of relying on line memories to store and process image data, the system uses software-based compensation curves that achieve the same dynamic range adjustment with minimal hardware resources, substituting hardware complexity with computational efficiency.
Data Source
AI summary
An apparatus for processing an image in a digital camera includes an image signal collection unit configured to process an image collected from a lens into image information using a CMOS image sensor (CIS), and an image correction unit configured to compensate for an intensity in response to a compensation curve corresponding to the image information collected by the image signal collection unit and output an image signal compensated depending on the intensity.