3D HOG Feature Calculation for Target Cell Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current image processing techniques face challenges in accurately detecting target cells, particularly nucleated red blood cells, when the shape in the thickness direction is not effectively utilized.
Innovation Solution
An image processing device that acquires images at multiple focal distances, selects the most focused image, sets an object region, calculates a three-dimensional histogram of oriented gradients (HOG) feature quantity, and uses machine learning to determine the presence of target cells based on defined conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If image processing is performed using only two-dimensional in-plane features, then the processing is simple and fast, but the detection accuracy of target cells is insufficient
Solution Approach 1:
The patent extends the traditional two-dimensional HOG feature calculation to three dimensions by incorporating the focal distance direction. The calculation unit computes luminance gradients not only in the in-plane direction but also in the focal distance direction, creating a three-dimensional HOG feature quantity that captures cellular shape information across different depths. This dimensional extension enables the system to detect target cells with higher accuracy by utilizing depth information without significantly increasing computational complexity.
2Measurement precision
If multiple captured images at different focal distances are processed, then detection accuracy improves, but the amount of data and processing time increase
Solution Approach 1:
The patent merges multiple captured images at different focal distances into a unified three-dimensional HOG feature quantity calculation. Instead of processing each image separately and then comparing results, the system integrates information from all focal planes simultaneously by calculating luminance gradients in the focal distance direction. This merging approach allows the determination unit to make more accurate decisions about target cell presence by considering depth information across all images in a single processing step, thereby reducing overall processing time while maintaining high detection accuracy.
Data Source
Figure 1
Figure 2
Figure 3A~3D
AI summary
An image processing device 20 acquires a plurality of captured images obtained by imaging a sample including a target cell at a plurality of different focal distances, sets an object region for detecting the target cell in at least two of the plurality of captured images, and performs determination about whether or not the target cell is included in the object region based on whether or not an image feature quantity based on change in image information in an image in-plane direction and a focal distance direction of an image in the object region in the plurality of captured images satisfies conditions defined for the target cell.