Urine Sediment Image Block Processing via Grid Feature Merging

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current urine sediment image processing methods face challenges in precision, particularly in distinguishing between different elements and background blocks, due to limited contextual information and sensitivity to small changes in input blocks.

Innovation Solution

The method involves dividing a block into grids, calculating n-dimensional local feature vectors, merging adjacent grids into intermediate blocks, and combining these feature vectors to create a general combination feature vector, which enhances processing precision by capturing contextual relationships between grids.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If a block to be processed is directly classified or retrieved using traditional machine learning methods, then the processing speed is maintained, but the precision is insufficient due to limited contextual information and sensitivity to small changes

Engineering Contradiction:
Improveprocessing precisionVSAvoidfeature extraction complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent divides a block to be processed into multiple grids, calculating n-dimensional local feature vectors for each grid. This segmentation allows the system to capture detailed local characteristics while maintaining overall context, thereby improving processing precision without excessively increasing complexity through hierarchical feature organization.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces multi-scale analysis by dividing blocks into grids at different scales and computing feature vectors that capture spatial relationships across multiple dimensions. This dimensional approach enriches contextual information, enabling more accurate classification and retrieval while systematically managing complexity through structured feature hierarchies.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Loss of information

If traditional feature extraction methods are used, then the computational complexity is low, but the contextual information captured is insufficient leading to reduced precision

Engineering Contradiction:
Improvecontextual informationVSAvoidprocessing complexity
Core Design Contradiction:
Loss of informationVSDevice complexity

Solution Approach 1:

The patent merges local feature vectors from multiple grids into comprehensive block-level features by calculating intermediate block merging feature vectors and combining them into general combination feature vectors. This merging process preserves contextual information across different spatial regions while organizing complexity through systematic combination rules.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent implements a nested feature hierarchy where local grid features are contained within intermediate block features, which are in turn contained within general block features. This nested structure preserves contextual information at multiple levels while managing complexity through hierarchical organization, allowing efficient information aggregation from fine-grained to coarse-grained representations.

Inventive Principle:
Principle #7Nested doll (Nesting)

Data Source

PatentEP3137894B1Method and apparatus for processing block to be processed of urine sediment image
Publication Date: 2019.04.10 SIEMENS HEALTHCARE DIAGNOSTICS INC
  • EP3137894B1 patent drawingFigure 1
  • EP3137894B1 patent drawingFigure 2
  • EP3137894B1 patent drawingFigure 3a~3d

AI summary

In the present invention are a method and apparatus for processing a block to be processed of a urine sediment image. The method comprises: dividing a block to be processed into a plurality of grids; calculating an n-dimensional local feature vector of each grid of the plurality of grids, where n is a positive integer; in the block to be processed, merging at least two adjacent grids of the plurality of grids into an intermediate block; calculating an intermediate block merging feature vector of the intermediate block; according to a predetermined combination rule, combining the intermediate block merging feature vectors obtained for different intermediate blocks of the block to be processed into a general combination feature vector of the block to be processed; and by way of taking the general combination feature vector as a feature in a feature set of block processing, processing the block to be processed.