HDR Image Processing Device Reducing Subject Blur via Region Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image processing devices that perform High Dynamic Range (HDR) composition struggle to reduce subject blur effectively, requiring complex control processes and influencing subsequent image processing steps, especially when using image information from moving subjects.

Innovation Solution

An image processing device that differentiates image processing between regions with moving subjects and other regions, using region data to selectively apply pixel values and exposure adjustments, thereby reducing subject blur and improving subsequent processing steps.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If HDR composition is performed using image information from moving subjects only, then subject blur is reduced, but the influence of processing during merging becomes significant in subsequent steps

Engineering Contradiction:
Improvesubject blur reductionVSAvoidquality of subsequent processing
Core Design Contradiction:
Manufacturing precisionVSReliability

Solution Approach 1:

The patent segments the image into multiple regions based on motion detection: regions with moving subjects and regions without moving subjects. Different image processing methods are applied to each segment. For regions with moving subjects, the patent uses image information from a single exposure to avoid subject blur, while for other regions, it performs normal HDR merging processing. This segmentation allows the system to reduce subject blur in critical areas without compromising the quality of subsequent processing steps across the entire image.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies different processing qualities to different regions of the image. In regions containing moving subjects, it uses high-quality single-exposure image information to maintain sharpness and avoid blur. In regions without moving subjects, it performs standard HDR composition to maximize dynamic range benefits. This local quality approach ensures that each region receives the most appropriate processing method for its characteristics, resolving the contradiction between subject blur reduction and overall processing quality.

Inventive Principle:
Principle #3Local quality

2Manufacturing precision

If control processes are changed according to the presence of a subject, then subject blur is reduced, but the control process becomes complicated

Engineering Contradiction:
Improvesubject blur reductionVSAvoidcontrol process complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent performs preliminary motion detection and region classification before the main HDR composition process. By detecting moving subjects in advance and marking their regions, the system establishes region data that guides subsequent processing. This preliminary action simplifies the overall control process by avoiding the need for complex real-time adjustments during merging, as the processing strategy is predetermined based on the detected motion patterns.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system uses the image data itself to determine the processing strategy. By analyzing the input images for motion characteristics and automatically generating region data based on detected moving subjects, the system makes the control process self-regulating. This eliminates the need for external complex control mechanisms, as the images provide the necessary information to guide their own processing through the generated region data.

Inventive Principle:
Principle #25Self-service

Data Source

PatentUS9626750B2Image processing method for a composite image and image processing device, image processing program, and recording medium therefor
Publication Date: 2017.04.18 MORPHO INC
  • US9626750B2 patent drawing
  • US9626750B2 patent drawing
  • US9626750B2 patent drawing

AI summary

An image processing device that performs image processing on a composite image obtained by merging a first image and a second image having different exposure conditions. A data acquiring unit acquires the composite image and region data indicating a target region of the composite image. A image processing unit performs different image processes between the target region and a region other than the target region based on the region data. The images include a region on which a moving subject is rendered. In the region on which the moving subject is rendered, the composite image is generated by using a pixel value of one of the first image and the second image. The region data represents one or a plurality of regions generated based on the region of the composite image, on which the moving subject is rendered, as the target region.