Method for identifying siderophore-producing microorganisms based on visual imaging

Abstract

The invention discloses a method for identifying siderophore microorganisms based on visual imaging, which is a strategic method and comprises the following steps of: quickly acquiring identification results of siderophore functions of microorganisms in a CAS culture medium through a computer vision technology, intelligently arranging related identification data tables and outputting Excel files; the workload of related workers is greatly reduced, and the situations of misjudgment of experimental results and the like caused by personal empirical are reduced.

Description

Technical Field

[0001] The present invention relates to the fields of artificial intelligence and bio-organic technology, and in particular to a method for identifying iron carrier-producing microorganisms based on visual imaging. Background Art

[0002] Probiotics can promote the growth, development and stress resistance of plants, animals and even humans, and inhibit the growth of harmful microorganisms. Iron carriers are a type of iron-containing substances produced by microorganisms under iron deficiency conditions. 3+ Chelate with high affinity can promote the insoluble Fe in the environment 3+ Converted into soluble Fe 2+ , which promotes plant growth. The production of iron carriers is also an important indicator and identification work in the screening of probiotics for agricultural use. Therefore, the intelligent identification of iron carrier function identification and other directions in microbial culture experiments is of great significance.

[0003] Traditionally, after the microorganisms are cultured in the universal iron carrier identification medium CAS, the identification of the iron carrier production function of the microorganism mainly relies on the naked eye to identify the dissolution circle in the CAS medium and the manual measurement and recording of data with a ruler. This method is time-consuming and labor-intensive, and greatly increases labor costs. In recent years, with the development of technology, automated identification has gradually been widely used. However, the existing automated identification cannot solve some problems in the actual application process. Some methods cannot clearly draw the characteristic diagram of the microorganism, and there will be missed judgments in the unclear parts of the microorganism image. At the same time, it cannot meet the problems related to biostatistics.

[0004] With the rapid development of artificial intelligence technology, many industries need to transform into intelligent and automated ones. Artificial intelligence technology can well assist R&D personnel to complete repetitive work in the biological field. Through the image recognition technology in the field of artificial intelligence, the accuracy of microbial cultivation in applications can be improved, the identification information can be automatically generated into spreadsheets, and the relevant data can be batch-entered into the information system to facilitate batch management of data information. When faced with the work of identifying the growth of microorganisms in culture media in large quantities, if traditional manual recognition is used to judge, sort and record data, the burr phenomenon generated during the sorting process will affect the recognition effect of researchers and cause errors in statistical data. The process of recording data will take a lot of time and energy, which causes a great waste of human resources. In the face of large amounts of experimental data, it is reasonable to adopt intelligent recognition and automatic data recording methods. The relevant staff only need to upload the results to the computer in the form of photos, identify the image data of each group of microorganisms in the culture medium through image recognition technology, and then rely on the intelligent information management system to automatically generate and record experimental data.

[0005] Therefore, a new automated identification and data recording method is needed. Summary of the invention

[0006] In view of the fact that some methods of the existing automated identification technology cannot clearly draw the characteristic diagram of microorganisms, and the phenomenon of missed judgment will occur in the unclear parts of the microorganism image, which cannot meet the problem of biostatistics. The present invention provides a method for identifying iron-carrier-producing microorganisms based on visual imaging. It is a strategic method. Through computer vision technology, the identification results of the iron-carrier-producing function of microorganisms in CAS culture medium can be quickly collected, and the relevant identification data tables can be intelligently organized and output as Excel files, which greatly reduces the workload of relevant staff and reduces the misjudgment of experimental results caused by personal empiricism.

[0007] The technical principles of the present invention are as follows:

[0008] A method and system for sorting microorganisms in culture medium based on visual imaging. The main process is as follows:

[0009] 1) Preprocessing module: Store the microbial culture medium in the form of images and transfer it into the system as json data;

[0010] 2) Visual imaging module: The image of the microbial culture medium is input through the image grayscale algorithm, image binarization algorithm, contour scanning algorithm and feature value extraction to obtain the image of the microbial culture medium and the coordinates of the center of the circle;

[0011] 3) Data acquisition module: The microbial culture medium visual imaging module obtains the center coordinates of each culture medium and the culture medium data, and by establishing a radius matrix, it is transferred to an Excel table to record the experimental data, realizing automatic identification and establishing a microbial culture medium database.

[0012] The purpose of the present invention is achieved through the following technical solutions:

[0013] A method for identifying siderophore-producing microorganisms based on visual imaging comprises the following steps:

[0014] Step 1, pre-processing module: record the relevant microorganisms cultured in CAS culture medium in the form of photos, try to fill the entire picture with culture medium for computer recognition, and transfer json data to the microbial culture medium visual imaging module;

[0015] Step 2: Visual imaging module: The main function is to accurately detect and locate the location information of microorganisms in each image through computer vision algorithms, and perform precise sorting. The specific steps are as follows:

[0016] (1) Image preprocessing: Through the image grayscale algorithm of computer vision, the values ​​of the red, green and blue channels in the culture medium image are eliminated, and only the brightness information in the culture medium image is retained, making the image features easier to analyze and identify;

[0017] Through the image graying algorithm, the gray value of the culture medium image is divided into (0, 255). Based on the template image as the standard, domain experts define the gray value range of (173, 178) as the microorganism with the ability to produce siderophores;

[0018] (2) Feature extraction: Because the culture medium of microorganisms is a relatively small target object in the image, usually only occupies a very small part of the image, and the image features of microorganisms in each experiment are different, the present invention targets this feature:

[0019] An image binarization algorithm was used. The implementation process of the image binarization algorithm is as follows: experts identify that the grayscale value in the grayscale image is in the interval of (173, 178) as a sign of successful culture of microbial culture medium, and all the pixels with grayscale values ​​less than 173 in the grayed culture medium image are recorded as 0, and all the pixels with grayscale values ​​greater than 173 are recorded as 255;

[0020] Through the contour scanning algorithm, the inner diameter of the culture medium image of successful microbial culture can be obtained and recorded. Similarly, the outer diameter of the culture medium image is scanned with 178 as the threshold value.

[0021] (3) Target detection: Using the contour detection algorithm, the center coordinates of each culture medium are recorded, and a center matrix is ​​established on the culture medium to record the specific coordinates of the culture medium;

[0022] Step 3, data acquisition module: according to the contour scanning algorithm in the microbial culture medium visual imaging module in step 2, draw the characteristic image of each microbial culture medium, calculate the image area, record its center coordinates, and then draw the center coordinates of each culture medium by traversing the contour scanning algorithm to establish a center matrix, create a table to pass the data of each culture medium of the microbial culture medium, realize automatic identification and establish a microbial culture medium database.

[0023] Compared with the prior art, the present invention has the following advantages:

[0024] The method for identifying iron-carrier-producing microorganisms based on visual imaging described in the present invention is a system and method for sorting microbial culture medium samples to address the shortcomings of the prior art. The present invention uses a picture of a microbial culture medium as a basic sample. The visual imaging system is used to identify the position of the microbial culture medium in the culture medium through computer vision and image recognition technology, draw an image of the microbial culture medium through contour scanning, and record the center and radius information of the microbial image. The data acquisition system establishes a center coordinate system through the culture medium information and the basic information of each culture medium, generates an Excel table for the experimental data in the entire culture medium, realizes intelligent target recognition of microbial experiments and automatic information entry functions, saves manpower and time, and improves work efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 is a general flow chart of a method for identifying siderophore-producing microorganisms based on visual imaging of the present invention;

[0026] Figure 2 It is a schematic diagram of a visual imaging module for identifying siderophore-producing microorganisms based on visual imaging according to the present invention;

[0027] Figure 3 This is a schematic diagram of a data acquisition module for identifying siderophore-producing microorganisms based on visual imaging according to the present invention;

[0028] Figure 4 It is a schematic diagram of a system module for identifying iron carrier-producing microorganisms based on visual imaging according to the present invention. DETAILED DESCRIPTION

[0029] The specific implementation modes of the present invention are further described below in conjunction with examples.

[0030] Example:

[0031] A method for identifying siderophore-producing microorganisms based on visual imaging comprises the following steps:

[0032] When the iron carrier function needs to be identified and the information needs to be recorded on a large scale for probiotics, the iron carrier identification system of this solution can be used to draw microbial characteristic images and record their characteristic values. The whole process is intelligent and automated, and the main steps are as follows:

[0033] Step 1: Preprocessing module, record the experimental culture medium in the form of pictures, try to fill the whole picture with culture medium for computer recognition, and use JSON data to transfer to the microbial culture medium visual imaging module;

[0034] Step 2: The microbial culture medium visual imaging module is mainly used to accurately detect and locate the location information of each microbial culture medium through computer vision algorithms for precise sorting. The specific steps are as follows:

[0035] (1) Image preprocessing: Through the image grayscale algorithm of computer vision, the values ​​of the red, green and blue channels in the culture medium image are eliminated, and only the brightness information in the culture medium image is retained, making the image features easier to analyze and identify. Through the image grayscale algorithm, the grayscale value of the culture medium image is divided into (0, 255). Based on the characteristics of microorganisms, domain experts distinguish that the experiment is successful if the microorganism is in the grayscale range of (173, 178);

[0036] (2) Feature extraction, because the culture medium of microorganisms is a relatively small target object in the image, usually only occupying a very small part of the image, and the image features of the microorganisms in each experiment are different; in view of this feature, the present invention adopts an image binarization algorithm, and the implementation process of the image binarization algorithm is as follows: experts distinguish that the grayscale value in the grayscale image is in the interval of (173, 178) as a mark of successful culture of the microorganism culture medium. Then all the pixels in the grayscale image of the culture medium with a grayscale value less than 173 are recorded as 0, and all the pixels with a grayscale value greater than 173 are recorded as 255; the inner circle diameter of the culture medium image of successful microbial culture can be obtained through the contour scanning algorithm and recorded, and similarly, the outer circle diameter of the culture medium image is scanned with 178 as the threshold;

[0037] (3) Target detection: using a contour detection algorithm, the center coordinates of each culture medium are recorded, and a center matrix is ​​established on the culture medium to record the specific coordinates of the culture medium;

[0038] Step three: The data acquisition module draws the characteristic image of each microbial culture medium according to the contour scanning algorithm in the microbial culture medium visual imaging module in step two, calculates the image area, records its center coordinates, and then draws the center coordinates of each culture medium by traversing the contour scanning algorithm to establish a center matrix, creates a table to pass the culture medium data of each microbial culture medium, and realizes automatic identification and establishes a microbial culture medium database.

[0039] The above is an explanation of the embodiments of the present invention. However, the present invention is not limited to the above embodiments. Any changes within the spirit and principle of the present invention and without departing from the conceptual framework of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for identifying siderophore-producing microorganisms based on visual imaging, characterized in that: The following steps are involved: Step 1, pre-processing module: record the relevant microorganisms cultured in CAS culture medium in the form of photos, try to fill the entire picture with culture medium for computer recognition, and transfer json data to the microbial culture medium visual imaging module; Step 2: Visual imaging module: The main function is to accurately detect and locate the location information of microorganisms in each image through computer vision algorithms, and perform precise sorting. The specific steps are as follows: (1) Image preprocessing: Through the image grayscale algorithm of computer vision, the values ​​of the red, green and blue channels in the culture medium image are eliminated, and only the brightness information in the culture medium image is retained, making the image features easier to analyze and identify; Through the image graying algorithm, the gray value of the culture medium image is divided into (0, 255). Based on the template image as the standard, domain experts define the gray value range of (173, 178) as the microorganism with the ability to produce siderophores; (2) Feature extraction: An image binarization algorithm was used. The implementation process of the image binarization algorithm is as follows: experts identify the grayscale value in the interval (173, 178) of the grayscale image as a sign of successful microbial culture medium cultivation, and all pixels with grayscale values ​​less than 173 in the grayed culture medium image are recorded as 0, and all pixels with grayscale values ​​greater than 173 are recorded as 255; Through the contour scanning algorithm, the inner diameter of the culture medium image of successful microbial culture can be obtained and recorded. Similarly, the outer diameter of the culture medium image is scanned with 178 as the threshold value. (3) Target detection: Using the contour detection algorithm, the center coordinates of each culture medium are recorded, and a center matrix is ​​established on the culture medium to record the specific coordinates of the culture medium; Step 3, data acquisition module: according to the contour scanning algorithm in the microbial culture medium visual imaging module in step 2, draw the characteristic image of each microbial culture medium, calculate the image area, record its center coordinates, and then draw the center coordinates of each culture medium by traversing the contour scanning algorithm to establish a center matrix, create a table to pass the data of each culture medium of the microbial culture medium, realize automatic identification and establish a microbial culture medium database.

Images