Deep Learning Colony Enumeration for Faster Microorganism Detection

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing methods for quantifying colony forming units of microorganisms in food samples are complex, time-consuming, and prone to variability due to reliance on general-purpose computers and internet-based AI engines, which delay results and introduce inconsistencies.

Innovation Solution

A dedicated colony enumeration device using a single-board computer with integrated neural network circuitry and concurrent processing of image capture, enhancement, and transmission processes, enabling rapid identification of colony forming units through a pretrained deep learning model.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If general-purpose computers and internet-based AI engines are used for colony enumeration, then device complexity is reduced, but processing time increases and reliability decreases

Engineering Contradiction:

Improvedevice complexityVSAvoidprocessing time

Core Design Contradiction:

Device complexityVSLoss of time

Solution Approach 1:

The system segments the colony enumeration task into specialized functional modules: image capture unit, image enhancement unit, deep learning model unit, and interaction component. Each module is dedicated to a specific function, allowing optimized processing at each stage while maintaining overall system efficiency and reducing total processing time.

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The patent introduces an interaction component as an intermediary between the deep learning model and the user interface. This intermediary manages the projection of identification characteristics onto the combined image, coordinating data flow and processing timing to reduce overall system response time while maintaining complexity management.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Device complexity

If general-purpose computers are used for colony enumeration, then device complexity is reduced, but result consistency deteriorates

Engineering Contradiction:

Improvedevice complexityVSAvoidresult consistency

Core Design Contradiction:

Device complexityVSReliability

Solution Approach 1:

The system employs a self-contained architecture where the single-board computer with integrated neural network circuitry performs all processing functions independently. The deep learning model processes images through dedicated hardware, and the interaction component automatically projects identification characteristics without requiring external computational resources, ensuring consistent and reproducible results.

Inventive Principle:

Principle #25Self-service

3Device complexity

If sequential processing is used for image capture, enhancement, and transmission, then process simplicity is maintained, but productivity decreases

Engineering Contradiction:

Improveprocess simplicityVSAvoidprocessing speed

Core Design Contradiction:

Device complexityVSProductivity

Solution Approach 1:

The system implements continuous processing where image capture, enhancement, and transmission operations overlap in time. The image capture unit continuously captures images while the image enhancement unit simultaneously processes previous images, and the deep learning model continuously identifies colony forming units, maintaining productive action throughout the entire workflow without idle periods.

Inventive Principle:

Principle #20Continuity of useful action

Solution Approach 2:

The image enhancement unit performs preliminary processing of captured images while the deep learning model is ready to receive and process them. This preliminary action includes enhancing image quality and preparing data structures before final processing, allowing the system to maintain high throughput by preparing multiple images in advance.

Inventive Principle:

Principle #10Preliminary action

4Device complexity

If deep learning model identification is performed without concurrent processing, then implementation simplicity is maintained, but time to result increases

Engineering Contradiction:

Improveimplementation simplicityVSAvoidtime to result

Core Design Contradiction:

Device complexityVSDuration of action of moving object

Solution Approach 1:

The system performs preliminary actions by capturing multiple images at different illumination settings and enhancing them in advance before the final identification is needed. The deep learning model receives pre-processed images, allowing it to quickly identify colony forming units without waiting for sequential processing of all images, thereby reducing time to result.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS12522798B2Microorganic detection system using a deep learning model

Publication Date: 2026.01.13 3M INNOVATIVE PROPERTIES CO

AI summary

Aspects of the present disclosure relate to a method of colony enumeration. The method includes identifying colony forming units of microorganisms in a combined image using a pretrained deep learning model on a colony enumeration device. The method can include providing a plurality of identification characteristics of the colony forming units to an interaction component such that the interaction component can project at least some of the plurality of identification characteristics onto the combined image.