3D Image Processing Apparatus Noise Reduction via Search Area Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing image processing techniques face inefficiencies in calculating the voxel with closest luminance from the neighborhood of a corresponding voxel, leading to noise in subtraction images due to errors in registration or intensity differences.
Innovation Solution
An image processing apparatus sets a search area around the corresponding position in the second image, using a shift of voxel position to determine the search area size based on voxel sizes of both images, and compares intensities within this area to obtain a difference value, reducing noise in three-dimensional subtraction images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If a voxel-matching method is used to reduce noise in subtraction images, then noise reduction is achieved, but calculation efficiency deteriorates due to the complexity of finding the voxel with closest luminance from the neighborhood
Solution Approach 1:
The patent divides the search area into multiple regions and processes voxels in segments rather than exhaustively searching the entire neighborhood. This segmentation approach reduces the computational complexity while maintaining effective noise reduction by focusing on the most relevant voxel candidates.
Solution Approach 2:
Instead of examining all voxels in the neighborhood (excessive action), the patent applies a threshold-based selection to consider only partial voxels that meet certain luminance criteria. This partial action approach maintains sufficient noise reduction effectiveness while significantly improving calculation efficiency.
2Reliability
If the search area is enlarged to find the closest luminance voxel, then noise reduction improves, but calculation time increases
Solution Approach 1:
The patent performs preliminary sorting or pre-processing of voxel luminance values before the actual search. By preparing the data structure in advance (e.g., sorting voxels by luminance or creating a hierarchical representation), the system can quickly identify the closest luminance match without examining all voxels, thus reducing calculation time while maintaining an adequate search area for noise reduction.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An apparatus that can obtain a subtraction image efficiently is provided. An image processing apparatus obtains a target image constituted by a set of voxels arranged in a discretized manner; sets a search area in the target image; and obtains, in a partial area included in the search area, on the basis of at least one of a voxel value included in the partial area or an interpolated value obtained by interpolation of a voxel of the target image, at least one of a maximum and a minimum of a voxel value or an interpolated value within the search area.