Infrared Component Removal Noise Reduction in Image Processing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional image processing techniques for document authentication, such as those described in JP-2007-43427-A and US 2008/00283729 A1, suffer from increased noise in visible images after removing the infrared component, leading to degraded image quality.

Innovation Solution

An image processing apparatus and method that includes a noise reduction process using linear or non-linear filters prior to infrared component removal, employing a noise reducer and a removal calculator to minimize noise and maintain image quality by subtracting infrared signals from mixed image signals with appropriate coefficients, and optionally using an offsetter to adjust pixel values.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If infrared component removal is performed by subtracting infrared image from visible image, then infrared component is successfully removed from visible image, but noise increases and image quality degrades

Engineering Contradiction:
Improveinfrared component removal accuracyVSAvoidimage quality
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent applies noise reduction processing to both the visible image and infrared image before performing the subtraction operation. This preliminary noise reduction step prepares the images by removing noise components in advance, so that when the infrared component is removed by subtraction, the resulting visible image does not have excessively increased noise. The noise reduction is performed as a preliminary action before the main infrared removal operation.

Inventive Principle:
Principle #10Preliminary action

2Reliability

If noise reduction is applied to reduce noise in visible image, then image quality is improved, but infrared component removal accuracy may be compromised

Engineering Contradiction:
Improveimage qualityVSAvoidinfrared component removal accuracy
Core Design Contradiction:
ReliabilityVSMeasurement precision

Solution Approach 1:

The patent creates a copy of the infrared image and uses it as a reference for noise reduction processing on the visible image. By copying the infrared image characteristics and applying them to the noise reduction process, the method can reduce noise in the visible image while maintaining accuracy in infrared component removal. The copied infrared image data serves as a guide for selective noise reduction.

Inventive Principle:
Principle #26Copying

3Reliability

If strong noise reduction filtering is applied, then noise is significantly reduced, but image resolution and detail are lost

Engineering Contradiction:
Improvenoise levelVSAvoidimage resolution
Core Design Contradiction:
ReliabilityVSManufacturing precision

Solution Approach 1:

The patent applies noise reduction processing with different characteristics to different parts of the image. By analyzing local image characteristics, the noise reduction is adapted to preserve edges and important details in regions where they are present, while still reducing noise in uniform regions. This local adaptation ensures that noise reduction does not uniformly blur the entire image, thereby maintaining resolution and detail where needed.

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP3767529B1Image processing apparatus, image processing method, and carrier medium
Publication Date: 2025.01.15 RICOH CO LTD
  • EP3767529B1 patent drawingFigure 1~2
  • EP3767529B1 patent drawingFigure 3
  • EP3767529B1 patent drawingFigure 4

AI summary

An image processing apparatus (100) includes a light source (2), an image sensor (9), and an invisible component remover (223). The light source emits visible component light and invisible component light to a target object. The image sensor receives reflected light from the target object to detect a visible invisible mixture image including an invisible component and a visible component and an invisible image including an invisible component. The invisible component remover, based on the visible invisible mixture image and the invisible image which have been detected, removes the invisible component from the visible invisible mixture image to generate a visible image. The invisible component remover includes a removal calculator (226) performing removal calculation of an invisible component with respect to the visible invisible mixture image, and a noise reducer (225) performing noise reduction on at least one of images to be input to and output from the removal calculator.