Density Filter Optical Low Pass Filtering for Imaging Noise Reduction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional imaging apparatuses face long processing times due to the need for extensive low pass filter operations on large numbers of pixels, which is inefficient and costly, especially in real-time applications like bar code reading and living body authentication.
Innovation Solution
Incorporating a density filter with varying transmittance characteristics between the lens and image sensor, allowing for optical low pass filtering by adjusting the distance between the lens and image sensor, thereby spreading the image over multiple pixels and substituting the need for conventional low pass filter operations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If low pass filter operations are performed on all pixels to remove noise, then noise removal is achieved, but processing time becomes excessively long
Solution Approach 1:
The patent applies preliminary action by performing low pass filter processing on only a selected portion of pixels (e.g., every nth pixel or pixels at specific locations) rather than all pixels. This preliminary filtering on a subset provides sufficient noise removal for the entire image while dramatically reducing processing time, as the filtered pixels serve as reference points for reconstructing or inferring information about unfiltered pixels.
2Reliability
If extensive low pass filter operations are performed on large numbers of pixels, then noise is removed effectively, but processing cost increases
Solution Approach 1:
The patent applies partial action by performing low pass filter processing on only a partial set of pixels rather than all pixels. By selecting a representative subset of pixels for filtering (such as pixels at regular intervals or pixels in critical regions), the system achieves adequate noise removal for the entire image while significantly reducing computational resources and processing costs.
3Loss of information
If clear images are obtained by matching image forming surface to sensor surface, then information volume is maximized, but high frequency noise is superposed on the signal
Solution Approach 1:
The patent applies local quality by applying low pass filter processing selectively to specific pixels rather than uniformly to all pixels. By identifying and filtering only certain pixels (such as pixels likely to contain noise or pixels at regular intervals), the system maintains high information volume from clear imaging while removing high frequency noise from critical regions, achieving a balance between information preservation and noise removal.
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
This approach significantly reduces image processing time, enabling the use of less expensive CPUs and reducing costs while maintaining real-time processing capabilities for imaging applications.
Implementation Method 1
a filter arranged at an aperture of the iris or near the aperture and having the transmittance varying with the position... The distance between the above mentioned lens and the image sensor is adjusted so that the image forming position can be set in front of or backward the light-receptive surface of the image sensor, thereby spreading the light assigned a desired transmittance characteristic by the filter within a predetermined range.
Data Source
AI summary
The optical path of the light having spread at each point of an object to be shot is limited by an iris unit of a density filter. Since the density filter unit has a desired transmittance characteristic, the amount of the light passing through the density filter unit continuously changes from the center of the luminous flux toward the periphery. After the luminous flux forms an image on the image forming surface, it spreads toward the light-receptive surface of an image sensor, and becomes an image spreading on the light-receptive surface in a predetermined range.