Weighted Binning Image Sensor for Resolution and Power Trade-offs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional binning techniques in image sensors reduce image resolution and dynamic range, leading to jagged edges and increased power consumption, while attempts to correct these issues either degrade resolution or fail to effectively address the problem.
Innovation Solution
An image sensor employing weight-based binning with vertical and horizontal interpolation units, along with a linear correction unit, to generate high-resolution images by setting weights for each row and column region based on illuminance information and exposure time, and applying these weights to pixel values to maintain or correct pixel values accordingly.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Use of energy by moving object
If conventional binning technique is applied to improve sensitivity and reduce power consumption, then power consumption is reduced and sensitivity is improved, but image resolution is degraded and jagged edges occur
Solution Approach 1:
The patent applies different binning weights to different spatial regions of the image sensor. Specifically, central pixels are given higher weights than peripheral pixels during the binning operation. This local differentiation allows the system to maintain high resolution in the central region where detail is most important, while still achieving sensitivity improvement and power reduction in peripheral regions through binning.
Solution Approach 2:
The patent changes the binning parameters by introducing variable weights instead of uniform binning. The weight values are adjusted based on spatial position (central vs. peripheral regions) and can be dynamically modified. This parameter variation enables the system to optimize between resolution and sensitivity differently across the image, preventing the uniform resolution degradation that occurs with conventional binning.
2Reliability
If conventional binning technique is applied to improve sensitivity, then sensitivity is improved, but dynamic range is reduced
Solution Approach 1:
By applying different weights to different spatial regions, the patent preserves more information locally in central regions where high dynamic range is critical. The weighted binning operation maintains finer granularity in signal representation for central pixels, preventing the uniform information loss that occurs in conventional binning across the entire sensor array.
3Object-generated harmful factors
If correction filter is applied to reduce jagged edges, then jagged edge phenomenon is reduced, but image resolution is degraded
Solution Approach 1:
The patent performs weighted binning as a preliminary operation before demosaicing, with weights carefully designed to anticipate and prevent jagged edge formation. By adjusting the weights of pixels participating in binning operations, the system proactively smooths potential jagged edges during the binning stage itself, rather than requiring subsequent correction filters that would degrade resolution.
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 weight-based binning method prevents resolution degradation, expands dynamic range, and reduces jagged edges by generating high-quality images with improved resolution and power efficiency.
Implementation Method 1
a pixel unit configured to output a bayer pattern by converting an optical signal of a subject into an electrical signal
Data Source
AI summary
An image sensor includes a pixel unit configured to output a bayer pattern by converting an optical signal of a subject into an electrical signal; and a vertical interpolation unit configured to generate a vertical binning image by adding or averaging values of vertical pixels of the bayer pattern based on weight information set for each row region of the bayer pattern. Further, the image sensor includes a horizontal interpolation unit configured to generate a horizontal binning image by adding or averaging values of horizontal pixels of the bayer pattern based on weight information set for each column region of the bayer pattern; and an image composition unit configured to generate a weight-based binning image by composing the vertical and horizontal binning images.