Bacteria Imaging Inspection Using Luminance-Validated Sharpening
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods struggle to accurately detect bacteria in images where proliferation occurs and foreign matters are present, leading to difficulties in isolating and counting individual bacteria due to reduced luminance differences and overlapping contours.
Innovation Solution
An imaging and image-processing system that calculates statistics on luminance values, excludes regions with abnormal luminance, performs contrast enhancement, and validates the process based on statistical ratios to determine bacterial proliferation accurately.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If bacteria proliferation progresses and bacteria do not appear in isolation, then the number of bacteria increases exponentially, but it becomes difficult to separate and correctly detect individual bacteria due to overlapping contours
Solution Approach 1:
The patent applies segmentation by dividing the image processing into multiple stages: initial thresholding to separate bacteria from background, contour extraction to identify individual boundaries, and iterative refinement to separate adjacent bacteria. This multi-stage segmentation approach enables accurate detection even when bacteria are densely packed and contours overlap.
Solution Approach 2:
The patent transitions from 2D image analysis to 3D spatial reasoning by utilizing depth information and spatial relationships between detected contours. By considering the spatial arrangement and overlap patterns in multiple dimensions, the system can distinguish individual bacteria even when their 2D projections overlap significantly.
2Quantity of substance
If the number of bacteria in the image increases, then the dynamic range for accurate capture expands, but the difference in shading between background region and bacterium region is reduced
Solution Approach 1:
The patent performs preliminary contrast enhancement and histogram equalization before the main detection process. By pre-processing the image to maximize luminance differences between bacteria and background, the system prepares optimal conditions for subsequent thresholding and contour extraction, ensuring accurate detection across the full dynamic range of bacterial densities.
Solution Approach 2:
The patent dynamically adjusts processing parameters based on the detected bacterial density. When bacterial concentration is high and luminance differences are reduced, the system modifies threshold values, contrast enhancement strength, and detection sensitivity parameters to maintain optimal detection performance across varying conditions.
3Adaptability or versatility
If foreign matter is present in the image, then detection complexity increases, but it becomes difficult to distinguish foreign matter from actual bacteria
Solution Approach 1:
The patent applies local quality analysis by examining morphological features, texture patterns, and spatial distribution characteristics at local regions of the image. By analyzing these local properties, the system can distinguish between foreign matter and actual bacteria based on their distinct visual characteristics, even when both are present in the same field of view.
Solution Approach 2:
The patent implements feedback mechanisms where detection results are continuously validated and refined. The system uses statistical analysis of detected objects' properties and compares them against expected bacterial characteristics, providing feedback that allows iterative refinement of detection accuracy and discrimination against foreign matter.
Data Source
Figure 1
Figure 2
Figure 3~5
AI summary
The present invention comprises: an imaging unit that captures an image of the inside of a container that retains a liquid containing fine particles; and an image-processing unit. The image-processing unit calculates a first statistic related to the spread of luminance values in the image, excludes a region having abnormal luminance values in the image on the basis of the result of comparison of the luminance values in the image with a predetermined luminance threshold, performs a correction process of sharpening the image from which the region having abnormal luminance values has been excluded, calculates a second statistic related to the spread of luminance values in the image subjected to the correction process, determines the validity of the correction process on the basis of the result of comparison of the ratio of the first statistic and the second statistic with a predetermined statistical ratio threshold, determines the degree of proliferation of fine particles on the basis of a feature amount of the image for which the correction process has been determined to be valid, and outputs the result of determination of the degree of proliferation. As a result, it is possible to more accurately extract bacteria and detect a region in which bacteria corresponding to the degree of proliferation are present regardless of the difference in the degree of proliferation or the presence or absence of foreign matters other than the bacteria included in an image including a particulate object to be measured such as growing bacteria.