A method, device, equipment and medium for detecting fungi based on image recognition assistance

Abstract

The application discloses a kind of based on image recognition auxiliary fungal detection method, device, equipment and medium, the method includes: based on image acquisition equipment acquires the image to be detected of sample liquid to be detected;According to the angle of acquisition, the image to be detected is fused, and target fused image is obtained;Feature extraction is carried out to the fused image, and feature extraction result is obtained;Based on fungal morphological characteristics and color characteristics are matched in the first fungal feature database, and first target fungal feature is obtained, and based on the texture characteristics formed by characteristic metabolite dyeing are matched in the second fungal feature database, and second target fungal feature is obtained;Based on first target fungal feature, the kind of fungus in sample liquid to be detected is determined, and based on second target fungal feature, the survival state of fungus in sample liquid to be detected is determined.The application realizes accurate and fast detection and identification of fungus.

Description

Technical Field

[0001] This invention relates to the field of biotechnology, and in particular to a method, apparatus, device, and medium for fungal detection based on image recognition assistance. Background Technology

[0002] Fungal infections have become an increasingly serious problem in the fields of medical and biological research. With the widespread use of immunosuppressive therapy, organ transplantation, and the overuse of antibiotics, the incidence of fungal infections is showing a significant upward trend. Traditional methods for fungal detection include microscopic observation and manual examination.

[0003] Microscopic observation methods, including direct smear examination and histopathological section examination, can provide information quickly to some extent, but their sensitivity is low, and they are prone to missing diagnoses in samples with small numbers of fungi or atypical morphology. Furthermore, these methods often struggle to distinguish between live and dead fungi, failing to provide accurate evidence for evaluating treatment effectiveness. Manual observation involves collecting fungal distribution maps and manually comparing them with images of known normal and abnormal samples to detect and identify fungi. However, subjective interpretation and fatigue during manual slide reading can lead to misjudgments, resulting in inaccurate test results. Summary of the Invention

[0004] To address the aforementioned problems, this invention provides an image recognition-assisted fungal detection method, apparatus, device, and medium for accurate and rapid detection and identification of fungi.

[0005] In a first aspect, the present invention provides an image recognition-assisted fungal detection method, comprising:

[0006] The image to be tested is acquired using an image acquisition device; the image to be tested is obtained after the sample solution to be tested has been stained with a fungal triple fluorescent staining solution.

[0007] The images to be detected are fused according to the acquisition angle to obtain the target fused image;

[0008] Feature extraction is performed on the fused image to obtain feature extraction results; the feature extraction results include fungal morphological features, color features, and texture features formed by staining with characteristic metabolites;

[0009] The first target fungal feature is obtained by matching the morphological and color features of fungi in the first fungal feature database, and the second target fungal feature is obtained by matching the texture features formed by staining of characteristic metabolites in the second fungal feature database. The first fungal feature database contains standard features of various fungi under specific culture and staining conditions, and the second fungal feature database contains texture features formed by staining of metabolites of live fungi and texture features formed by staining of metabolites of dead fungi.

[0010] The species of fungi in the sample solution to be tested are determined based on the characteristics of the first target fungus, and the survival status of the fungi in the sample solution to be tested is determined based on the characteristics of the second target fungus.

[0011] Secondly, the present invention provides an image recognition-assisted fungal detection device, comprising:

[0012] The image acquisition module is used to acquire images of the sample solution to be tested based on the image acquisition device; the image to be tested is an image acquired after the sample solution to be tested has been stained with fungal triple fluorescent staining solution;

[0013] The image fusion module is used to fuse the images to be detected according to the acquisition angle to obtain the target fused image;

[0014] The feature extraction module is used to extract features from the fused image to obtain feature extraction results; the feature extraction results include fungal morphological features, color features, and texture features formed by staining with characteristic metabolites;

[0015] The feature matching module is used to match the first fungal feature database based on the morphological and color features of fungi to obtain the first target fungal feature, and to match the second fungal feature database based on the texture features formed by staining of characteristic metabolites to obtain the second target fungal feature. The first fungal feature database contains standard features of various fungi under specific culture and staining conditions, and the second fungal feature database contains texture features formed by staining of metabolites of live fungi and texture features formed by staining of metabolites of dead fungi.

[0016] The fungal detection module is used to determine the species of fungi in the sample solution to be tested based on the characteristics of the first target fungus, and to determine the survival status of the fungi in the sample solution to be tested based on the characteristics of the second target fungus.

[0017] Thirdly, the present invention also provides an electronic device, comprising: a memory for storing computer software programs; and a processor for reading and executing the computer software programs, thereby realizing the image recognition-assisted fungal detection method as described in the first aspect above.

[0018] Fourthly, the present invention also provides a non-transitory computer-readable storage medium storing a computer software program, which, when executed by a processor, implements the image recognition-assisted fungal detection method as described in the first aspect above.

[0019] Fifthly, the present invention provides a computer program product, including a computer program that, when executed by a processor, implements the image recognition-assisted fungal detection method as described in the first aspect above.

[0020] The image recognition-assisted fungal detection method provided by this invention acquires more comprehensive information about fungal samples through multi-angle, high-resolution image acquisition, avoiding missed diagnoses due to low fungal quantity or atypical morphology. During image preprocessing, segmentation, and feature extraction, feature extraction comprehensively assesses the fungal morphological characteristics, color features, and texture features formed by staining with characteristic metabolites. This accurately highlights fungal features, unlike microscopic observation which struggles to distinguish between live and dead fungi. Therefore, compared to manual observation, feature extraction and matching based on objective image recognition algorithms avoids subjectivity and fatigue-induced misjudgments in manual image reading, greatly improving the accuracy of detection results and enabling accurate and rapid detection and identification of fungi. Attached Figure Description

[0021] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only one embodiment of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0022] Figure 1 This is a schematic flowchart of the image recognition-assisted fungal detection method provided by the present invention;

[0023] Figure 2 This is a schematic diagram of the structure of the image recognition-assisted fungal detection device provided by the present invention;

[0024] Figure 3 This is a schematic diagram of the structure of the electronic device provided by the present invention;

[0025] Figure 4 This is a schematic diagram of an embodiment of the computer-readable storage medium provided in this invention. Detailed Implementation

[0026] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0027] In the description of this invention, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the stated features. In the description of this invention, "a plurality of" means two or more, unless otherwise explicitly specified. In the description of this invention, the term "for example" is used to mean "used as an example, illustration, or description." Any embodiment described as "for example" in this invention is not necessarily to be construed as being more preferred or advantageous than other embodiments. The following description is provided to enable any person skilled in the art to make and use the invention. In the following description, details are set forth for illustrative purposes. It should be understood that those skilled in the art will recognize that the invention can be made without using these specific details. In other instances, well-known structures and processes are not described in detail to avoid obscuring the description of the invention with unnecessary detail. Therefore, the invention is not intended to be limited to the embodiments shown, but is consistent with the broadest scope of the principles and features disclosed herein.

[0028] Optional, refer to Figure 1 , Figure 1 This is a flowchart illustrating the image recognition-assisted fungal detection method provided by the present invention. The executing entity of the image recognition-assisted fungal detection method provided by the present invention can be a fungal detection device, such as... Figure 1 As shown, image recognition-assisted fungal detection methods include the following:

[0029] Step 10: Acquire the image of the sample solution to be tested using an image acquisition device. The image to be tested is obtained after the sample solution has been stained with a fungal triple fluorescent staining solution.

[0030] Optionally, the image acquisition device in the fungal detection apparatus is used to acquire images of the sample solution to be tested. Prior to this, the sample solution must be stained with a triple fluorescent staining solution for fungi. This staining solution can produce a specific fluorescent reaction with the fungi in the sample, making the fungi easier to identify and analyze in the image. The image acquisition device must have appropriate resolution, sensitivity, and other parameters to ensure that the acquired images are clear and contain sufficient detail for subsequent processing and analysis.

[0031] In one embodiment, the fungal detection device is equipped with a high-resolution fluorescence microscope camera as an image acquisition device. First, the sample solution to be tested is prepared and placed on a glass slide, and a triple fluorescent staining solution for fungi is evenly added. After an appropriate staining time, the slide is placed on the stage of the fluorescence microscope. The fluorescent substances in the sample are excited to emit light through the microscope's optical path system. The camera is set to high-resolution mode (e.g., 5000*5000 pixels), and the sensitivity is adjusted to obtain a clear fluorescence image. At this time, the fungal detection device controls the camera to capture an image of the sample solution to be tested, obtaining an image containing the fungi to be tested. In the image, the fungi exhibit bright colors due to the fluorescent staining, forming a sharp contrast with the background.

[0032] Step 20: Fuse the images to be detected according to the acquisition angle to obtain the target fused image.

[0033] Furthermore, since images acquired from different angles may contain information about different aspects of fungi, these images need to be fused for a more comprehensive analysis of fungal characteristics. The fungal detection device processes multiple images according to their acquisition angles, and uses a specific image fusion algorithm to combine the useful information from these images to generate a single fused image. This fused image comprehensively reflects the morphological, color, and other features of fungi from various angles, providing richer data for subsequent feature extraction, as described in steps 201 to 204.

[0034] Continuing with the example above, the fungal detection device acquires four images from multiple angles (such as 0°, 30°, 60°, and 90°) by controlling the rotation of the stage of the fluorescence microscope. The device employs a weighted average fusion algorithm, which assigns different weights to each image based on factors such as sharpness and contrast. For example, the image at 0° has the highest sharpness, with a weight of 0.4; the images at 30° and 60° are next, each with a weight of 0.25; and the image at 90° is relatively poor, with a weight of 0.1. The algorithm calculates and averages the color values ​​of the corresponding pixels in these four images according to their weights, ultimately generating a fused image of the target.

[0035] Step 30: Feature extraction is performed on the fused image to obtain the feature extraction results. The feature extraction results include fungal morphological features, color features, and texture features formed by staining with characteristic metabolites.

[0036] Furthermore, the fungal detection device extracts features based on the characteristics of fungi, primarily including morphological features (such as shape, size, hyphal thickness and branching), color features (fungi exhibit specific colors due to fluorescent staining, and their hue, saturation, etc., can be used as features), and texture features formed by staining with characteristic metabolites (metabolic products react with the staining solution to form unique texture patterns on or around the fungus). In one embodiment, for morphological feature extraction, the fungal detection device uses an edge detection algorithm to identify the fungal outline, calculates the size of the fungus using parameters such as the perimeter and area of ​​the outline, and analyzes the shape of the outline to determine whether it is circular, elliptical, or irregular. For the hyphae, a thinning algorithm is used to thin the hyphae to a single pixel width. For color feature extraction, the fused image is converted from the RGB color space to the HSV color space, and the hue, saturation, and brightness values ​​of the fungal region are extracted as color features. For texture feature extraction, a gray-level co-occurrence matrix algorithm is used to calculate parameters such as contrast, correlation, energy, and entropy of the fungal region's texture. For example, the analysis revealed that the target fungus was elliptical in shape, with a major axis length of 20 μm and a minor axis length of 15 μm; in terms of color features, the hue value was 120, the saturation was 0.8, and the brightness was 0.6; in terms of texture features, the contrast was 0.5, the correlation was 0.6, the energy was 0.7, and the entropy was 0.4.

[0037] Step 40: Match the fungal morphological and color features in the first fungal feature database to obtain the first target fungal feature, and match the texture features formed by staining with characteristic metabolites in the second fungal feature database to obtain the second target fungal feature.

[0038] Furthermore, the fungal detection device performs feature matching in two different fungal feature databases. The first fungal feature database stores standard features of various fungi under specific culture and staining conditions. The device compares the extracted fungal morphological and color features with the features in this database, finds the most matching feature, and determines the first target fungal feature, as described in steps 401 to 404.

[0039] Furthermore, the second fungal feature database contains texture features formed by staining of metabolites of live and dead fungi. The device matches the extracted texture features in this database to determine the second target fungal feature, as described in steps 405 to 408.

[0040] Step 50: Determine the species of fungi in the sample solution to be tested based on the characteristics of the first target fungus, and determine the survival status of the fungi in the sample solution to be tested based on the characteristics of the second target fungus.

[0041] Furthermore, the fungal detection device determines the species of fungi in the sample solution to be tested based on the characteristics of the first target fungus, as described in steps 501 to 503. Further, the fungal detection device can determine the survival status of the fungi in the sample solution to be tested based on the characteristics of the second target fungus. Because the second fungal feature database distinguishes the staining texture characteristics of the metabolic products of live and dead bacteria, the conclusion of whether the fungus is alive or dead can be drawn based on the matching results, as described in steps 504 to 506.

[0042] This invention, through multi-angle, high-resolution image acquisition, can obtain more comprehensive information about fungal samples, avoiding missed diagnoses due to low fungal quantity or atypical morphology. During image preprocessing, segmentation, and feature extraction, feature extraction comprehensively assesses the fungal morphological characteristics, color features, and texture features formed by staining with characteristic metabolites. This accurately highlights fungal features, unlike microscopic observation which struggles to distinguish between live and dead fungi. Therefore, compared to manual observation, feature extraction and matching based on objective image recognition algorithms avoids subjectivity and fatigue-induced misjudgments in manual image reading, greatly improving the accuracy of detection results and enabling accurate and rapid detection and identification of fungi.

[0043] In one embodiment, steps 201 to 204 are described as follows:

[0044] Step 201: Based on the acquisition angle, each image to be detected is grouped according to a preset angle range to obtain multiple initial detection image groups.

[0045] Optionally, the fungal detection device categorizes each acquired image into preset angle ranges based on the acquisition angle. These preset angle ranges are pre-set according to the detection requirements and the necessity of multi-view observation of fungi. This grouping method allows images with similar acquisition angles to be grouped together. Continuing with the previous example of the fungal detection device using a fluorescence microscope camera to acquire images of the sample solution from multiple angles, for example, if the preset angle range set by the device is 30°, then 0°-30° forms one group, 30°-60° another, 60°-90° another, and so on. The camera acquired six images from angles of 0°, 15°, 35°, 45°, 65°, and 75°. Therefore, the images acquired at 0° and 15° will be assigned to the initial detection image group of 0°-30°; the images acquired at 35° and 45° will be assigned to the initial detection image group of 30°-60°; and the images acquired at 65° and 75° will be assigned to the initial detection image group of 60°-90°, thus obtaining multiple initial detection image groups.

[0046] Step 202: Arrange the images to be detected in each initial detection image group according to the size of the acquisition angle to obtain the updated detection image group.

[0047] Furthermore, for each initial detection image group, the fungal detection device rearranges the images within the group according to the ascending order of their acquisition angles. This allows the images to exhibit ordered angular variations, facilitating better utilization of the correlation between images from different angles during subsequent feature extraction and fusion processes. This results in a more accurate reflection of the true state of fungi under continuous angular changes. Taking the initial detection image group of 0°-30° as an example, the group contains images acquired at 0° and 15°. The fungal detection device identifies the acquisition angles of these two images and then arranges them in ascending order of angle, placing the 0° image first and the 15° image last, forming the updated detection image group. Similarly, for the initial detection image group of 30°-60°, if the group contains images acquired at 35° and 45°, the device will arrange the 35° image first and the 45° image last, completing the arrangement of images within each initial detection image group to obtain the updated detection image groups.

[0048] Step 203: For each image in the updated detection image group, extract the key feature points of each image to be detected based on the feature extraction algorithm.

[0049] Furthermore, within each updated image group, the fungal detection device employs a specific feature extraction algorithm to extract key feature points for each image to be detected within the group. These key feature points accurately characterize the important structural and textural information of fungi in the image, serving as a crucial basis for subsequent image fusion. Different feature extraction algorithms (such as SIFT, SURF, etc.) identify unique and stable points in the image as key feature points based on information such as image grayscale and gradient.

[0050] For example, the fungal detection device uses the SIFT (Scale Invariant Feature Transform) algorithm to extract key feature points. Taking images acquired at 35° and 45° within a detection image group updated from 30° to 60° as an example, the device runs the SIFT algorithm on the 35° acquired image. This algorithm first constructs an image pyramid and detects extreme points in different scale spaces. After a series of Gaussian difference operations, key point localization, and orientation assignment, key feature points of the fungal portion in the image are extracted, such as multiple key feature points identified at the edges of the fungus and at hyphal bifurcation points. Similarly, the SIFT algorithm is also applied to the 45° acquired image to extract corresponding key feature points. These key feature points reflect the unique structural and textural features of the fungus at different angles, providing crucial data for subsequent fusion operations.

[0051] Step 204: Based on the key feature points of each image to be detected in each updated detection image group, perform fusion to obtain the target fused image.

[0052] Furthermore, the fungal detection device fuses the key feature points of each image to be detected in each updated detection image group to obtain the target fused image, as described in steps 2041 to 2044.

[0053] This invention enables the reasonable grouping and orderly arrangement of images acquired from different angles, and achieves precise fusion based on key feature points. The resulting fused image integrates important structural and textural information of fungi from multiple angles. Compared to images from a single angle, the fused image more comprehensively and accurately presents the true morphology and characteristics of fungi, providing a data foundation for subsequent fungal species identification and survival status analysis, thus improving the accuracy and reliability of fungal detection.

[0054] In one embodiment, steps 2041 to 2044 are described as follows:

[0055] Step 2041: For each image in the updated detection image group, determine the correspondence between different images based on the location and descriptor of the key feature points in each image to be detected, and obtain the image matching result. The image matching result indicates that the feature points represent the same object or region in different images.

[0056] Optionally, for each updated group of detected images, the fungal detection device uses the location and descriptor of key feature points in each image to explore the correspondence between different images. The location information of key feature points indicates their coordinates in the image, while the descriptor is a quantitative description of the features of the area surrounding the feature point, such as the grayscale variation pattern around the feature point. By comparing the descriptors of key feature points in different images, highly similar feature point pairs are found. These feature point pairs represent the same object or region in different images, thus yielding the image matching result.

[0057] In one embodiment, taking the 30°-60° updated detection image group as an example, this group contains images acquired from 35° and 45°, and key feature points have been extracted. For example, the image acquired at 35° contains feature point A, with position coordinates (x1, y1), and a descriptor of a set of quantized values ​​D1, containing information such as the grayscale changes around the point; the image acquired at 45° contains feature point B, with position coordinates (x2, y2), and a descriptor of D2. The fungal detection device calculates the similarity between D1 and D2, for example, by using Euclidean distance to calculate their difference. If the distance is less than a set threshold, feature points A and B are considered to match. By comparing all key feature points in the two sets of images in this way, multiple matching feature point pairs are found, such as (A, B), (C, D), etc. These matching pairs constitute the image matching results, indicating that the matching feature points in different images represent the same part of the fungus.

[0058] Step 2042: Based on the image matching results, determine the overlapping regions between images within each updated detection image group. The overlapping regions represent the parts that are commonly covered in different images.

[0059] Furthermore, the fungal detection device determines the overlapping regions between images within each updated detection image group. Since the matched feature points represent the same object or region, by analyzing the positional distribution of these matched feature points in different images, the commonly covered parts in different images, i.e., the overlapping regions, can be inferred. Typically, the coordinate range of the matched feature points can be used to define the boundaries of the overlapping regions. Continuing with the example of the 30°-60° updated detection image group, we know the matched feature point pairs (A,B), (C,D), etc. Feature point A has coordinates of (x1,y1), B has coordinates of (x2,y2), C has coordinates of (x3,y3), D has coordinates of (x4,y4), etc. The fungal detection device finds the minimum and maximum coordinate values ​​of all matched feature points in the horizontal and vertical directions. For example, if the minimum horizontal coordinate is min_x and the maximum horizontal coordinate is max_x, and the minimum vertical coordinate is min_y and the maximum vertical coordinate is max_y, then in images captured at 35° and 45° angles, the rectangular region with (min_x, min_y) as the top-left vertex and (max_x, max_y) as the bottom-right vertex is the overlapping region. Within this overlapping region, both images capture the same part of the fungus, only with slight differences in perspective.

[0060] Step 2043: For images within the same updated detection image group, an image fusion algorithm is used to fuse the images to be detected in the overlapping area to generate a fused image within the group.

[0061] Furthermore, for images within the same updated detection image group, within a defined overlapping region, the fungal detection device uses an image fusion algorithm to fuse the images to be detected, generating a fused image within the group. The image fusion algorithm comprehensively considers the pixel information of different images within the overlapping region, generating an image that contains more information and more accurately reflects the characteristics of that part of the fungus. Common fusion algorithms, such as multi-resolution analysis-based fusion methods, process images at different resolutions and then merge the results.

[0062] Continuing with the example of the 30°-60° updated detection image group, a wavelet transform-based multi-resolution analysis and fusion algorithm is employed within the defined overlapping region. Wavelet decomposition is performed on portions of the images acquired at 35° and 45° within the overlapping region to obtain sub-band images of different frequencies. For low-frequency sub-bands, the average value of corresponding pixels from both sub-bands is taken; for high-frequency sub-bands, pixels corresponding to the larger absolute value of the coefficients are selected. Then, the processed sub-band images undergo inverse wavelet transform to obtain the fused image portion of the overlapping region. For non-overlapping regions, the corresponding portions of the original images are directly retained. The fused result of the overlapping region and the non-overlapping region are combined to generate the intra-group fused image of the 30°-60° updated detection image group.

[0063] Step 2044: For images between different updated detection image groups, fuse the images within each updated detection image group again to obtain the target fused image.

[0064] Furthermore, the fungal detection device further fuses the intra-group fused images from different updated detection image groups to obtain the target fused image. This step integrates information from different angular ranges represented by each updated detection image group, generating a comprehensive image reflecting the overall characteristics of the fungus. The fusion method is similar to intra-group image fusion, but the images involved come from updated detection image groups with different angular ranges. For example, intra-group fused images of updated detection image groups such as 0°-30°, 30°-60°, and 60°-90° have already been generated. The fungal detection device further determines the overlapping areas and matching relationships between these intra-group fused images, for example, by detecting image edge features. Then, in the overlapping areas, an algorithm similar to intra-group image fusion (such as a fusion algorithm based on multi-resolution analysis) is applied for fusion. For non-overlapping areas, reasonable stitching and integration are also performed. Finally, the target fused image is generated, which covers information about the fungus from multiple angular ranges from 0° to 90°, comprehensively and meticulously displaying the morphology, structure, and texture characteristics of the fungus.

[0065] This invention begins by determining the correspondence between key feature points in different images, gradually identifying overlapping areas and performing intra- and inter-group image fusion. The resulting fused image integrates detailed information about fungi from multiple angular intervals, thus providing a more comprehensive and accurate representation of the fungus's overall appearance, including its complex structure and texture. This provides richer and more accurate data support for subsequent image-based fungal species identification and survival status assessment, improving the accuracy and reliability of fungal detection.

[0066] In one embodiment, steps 401 to 404 are described as follows:

[0067] Step 401: Decompose the fungal morphological and color features according to different dimensions to obtain the decomposed morphological and color feature components.

[0068] Optionally, the fungal detection device decomposes the extracted fungal morphological and color features. For morphological features, it is decomposed according to different dimensions, such as separating shape features into size features and hyphal features. For color features, it is decomposed into different component dimensions such as hue, saturation, and brightness to obtain more targeted color feature components.

[0069] In one embodiment, the fungal morphological features previously extracted by the fungal detection device are elliptical in shape, with a major axis of 20 μm and a minor axis of 15 μm, and branched hyphae. The color features, in the HSV color space, have a hue value of 120, a saturation of 0.8, and a lightness of 0.6. The device decomposes the morphological features into: elliptical features in the shape dimension; features with a major axis of 20 μm and a minor axis of 15 μm in the size dimension; and branched hyphae features in the hyphal dimension. The color features are decomposed into a hue component of 120, a saturation component of 0.8, and a lightness component of 0.6.

[0070] Step 402: Based on the decomposed morphological features and according to the type and combination of basic geometric elements, a preliminary comparison is performed with the standard fungal morphological features of the same classification level in the first fungal feature database to obtain the morphological feature matching results. The classification levels include phylum, class, order, family, genus, and species.

[0071] Furthermore, the fungal detection device, based on the decomposed morphological characteristics, performs a preliminary comparison with the standard fungal morphological characteristics of the same classification level in the first fungal characteristic database, according to the type of basic geometric elements (such as circles, ellipses, straight lines, etc., representing basic elements of different fungal morphologies) and the combination method (such as the branching pattern of hyphae). The classification levels, from largest to smallest, include phylum, class, order, family, genus, and species.

[0072] Continuing with the above example, the fungal detection device compares the decomposed elliptical shape, 20μm major axis, 15μm minor axis, and branched hyphae characteristics with the standard fungal morphological characteristics at the "Phylum" level in the first fungal feature database. In the database, different classes of fungi under the "Phylum" have their own corresponding standard morphological descriptions. The device found that the basic geometric elements and combinations of some fungi under the "Ascomycetes" class were quite similar; these fungi have elliptical cell morphology and may have branched hyphae. Therefore, the morphological feature matching results indicate some fungal categories related to the "Ascomycetes" class, narrowing the subsequent matching scope to within the "Ascomycetes" class.

[0073] Step 403: Based on the color feature components, the color feature matching result is compared with the standard fungal color feature corresponding to the index position in the first fungal feature database to obtain the color feature matching result.

[0074] Furthermore, the fungal detection device locates the standard fungal color feature at the corresponding index position in the first fungal feature database, and then compares it with the previously decomposed color feature components. Because the morphological feature matching results have already limited a certain range of fungal categories, comparing color features within this range can more accurately determine the target fungal features. By comparing the differences between color feature components such as hue, saturation, and brightness and the standard values ​​in the database, the degree of color feature matching is determined.

[0075] Continuing with the above embodiment, since the morphological feature matching result in step 402 points to some fungi under the "Ascomycetes" class, the fungal detection device finds the standard color features corresponding to these fungi within the "Ascomycetes" class in the first fungal feature database. For example, the standard color features of one possible matching fungus in the database are hue 115-125, saturation 0.7-0.9, and lightness 0.5-0.7. Comparing the previously decomposed hue component 120, saturation component 0.8, and lightness component 0.6 with this standard color feature, it is found that all components are within the standard range, indicating a high degree of color feature matching. Thus, the color feature matching result is obtained, further confirming that the fungal category corresponding to the standard color feature is within the matching range.

[0076] Step 404: Determine the characteristics of the first target fungus based on the color feature matching results.

[0077] Further, the fungal detection device, based on the color feature matching results obtained in step 403 and combined with the previous morphological feature matching results, finally determines the first target fungal feature. If the color feature matching result and the morphological feature matching result point to the same type of fungus in the database, and the degree of matching is high, then the feature of that type of fungus can be identified as the first target fungal feature. This feature represents the standard fungal feature that is closest to the fungal feature in the sample to be tested under specific culture and staining conditions. After the previous steps, the morphological feature matching points to some fungi under the "Ascomycetes" class, and the color feature matching also highly matches the standard feature of a specific fungal species under the "Ascomycetes" class. For example, through comprehensive judgment, it is determined that the feature of the fungus to be tested best matches the standard feature of "Ascomycetes-Saccharales-Candidae-Candida-Candida albicans". Therefore, the fungal detection device determines the first target fungal feature as the standard feature of Candida albicans in the first fungal feature database, including its specific morphological and color feature descriptions.

[0078] In this embodiment of the invention, fungal characteristics are first decomposed, and then compared with a first fungal characteristic database in stages. Ultimately, the database can accurately select the first target fungal characteristic that best matches the morphological and color characteristics of the fungus to be detected. This provides a precise reference when determining the fungal species in the sample solution, thus enabling more efficient and accurate finding of matching objects from a large database, improving the accuracy and efficiency of fungal species identification.

[0079] In one embodiment, steps 405 to 408 are described as follows:

[0080] Step 405: Extract the primary and secondary direction information and structural information from the texture features formed by the staining of characteristic metabolites based on the image structure tensor algorithm.

[0081] Optionally, the fungal detection device employs an image structure tensor algorithm to process the texture features formed by staining with characteristic metabolites. This algorithm constructs a structure tensor matrix by calculating the gradient information of local regions of the image. From this matrix, the primary and secondary direction information of the texture features can be extracted, i.e., the main and secondary directions of the texture in the image, which reflects the directionality of the distribution of fungal metabolites. Simultaneously, structural information, such as the complexity and regularity of the texture, can also be obtained. In one embodiment, for example, the fungal detection device has acquired a texture image formed by staining with characteristic metabolites in the sample to be detected. The device applies the image structure tensor algorithm to this image. Within a small local region of the image, a structure tensor matrix is ​​constructed by calculating the gradient values ​​of pixels. For example, within a 3*3 pixel neighborhood, the structure tensor matrix is ​​obtained after calculation. Eigenvalue decomposition is performed on this matrix; the eigenvector direction corresponding to the larger eigenvalue is the primary direction of the texture, and the eigenvector direction corresponding to the smaller eigenvalue is the secondary direction. Simultaneously, based on some properties of the matrix, such as the proportion of eigenvalues, the structural information of the texture in this region can be evaluated, determining whether it is a regular or irregular texture, and the complexity of the texture. After processing the entire texture image, the primary and secondary directional information and structural information of the texture features were obtained.

[0082] Step 406: Traverse the hierarchical storage structure of the second fungal feature database and determine the relevant information in each storage layer. The second fungal feature database is stored separately according to the texture features of live and dead fungal metabolites, and within each storage area, it is hierarchically divided according to different fungal species and metabolite types.

[0083] Furthermore, the second fungal feature database has a specific hierarchical storage structure. The fungal detection device begins to traverse this database. First, the database is stored separately according to the texture features of live and dead fungal metabolites. Within each storage area, it is further subdivided according to different fungal species and metabolite types. During the traversal, the device sequentially determines relevant information in each storage layer, including standard descriptions of the texture features formed by the staining of metabolites of different fungal species in live or dead states. These descriptions include information such as primary and secondary orientation information and structural information extracted in step 405. In one embodiment, the traversal of the second fungal feature database begins by entering the live fungal metabolite texture feature storage area, which is further subdivided according to fungal species, such as the Candida albicans layer, the Aspergillus layer, etc. Within the Candida albicans layer, it is further subdivided according to different metabolite types. Enter the Candida albicans - a specific metabolite type layer, read the standard information stored in this layer about the texture features formed by staining of Candida albicans in its live state by this metabolite, including the standard value range of the primary and secondary directions of the texture, structural information description, etc., traverse each layer under the dead bacteria metabolite texture feature storage area to obtain the corresponding information.

[0084] Step 407: Determine the primary and secondary directional information and structural information in the texture features based on the similarity measurement algorithm, and evaluate the similarity between them and the relevant information in each storage layer.

[0085] Furthermore, the fungal detection device employs a similarity measurement algorithm to compare the primary and secondary directional information and structural information of the texture feature to be detected extracted in step 405 with the relevant information obtained from each storage layer of the second fungal feature database in step 406. The similarity measurement algorithm calculates the degree of similarity between the two based on set rules, obtaining a similarity evaluation result. For example, for primary and secondary directional information, the difference in directional angles can be calculated; for structural information, the quantized value of texture complexity can be compared, etc. By comprehensively considering these comparison results, an overall similarity evaluation score or level is obtained. For example, from step 405, the primary directional angle of the texture feature to be detected is 45°, the secondary directional angle is 135°, and the structural information shows a texture complexity of 0.6 (quantized value). In the second fungal feature database, the standard primary directional angle of a certain metabolite texture feature in the live state of Candida albicans is 40°-50°, the secondary directional angle is 130°-140°, and the structural information describes a complexity of 0.5-0.7. The fungal detection device employs a similarity measurement algorithm, calculating that the differences in the primary and secondary directional angles are within reasonable ranges, and the quantifiable values ​​for texture complexity are also similar. Through comprehensive algorithmic calculation, a high similarity assessment result is obtained between the storage layer and the texture features to be detected, recorded in a certain form (e.g., a score of 80 out of 100). The device performs this similarity assessment operation on all storage layers in the database.

[0086] Step 408: Determine the characteristics of the second target fungus based on the similarity assessment results.

[0087] Furthermore, based on the similarity assessment results obtained in step 407, the fungal detection device determines the second target fungal feature. It identifies the fungal feature corresponding to the storage layer with the highest similarity among all storage layer similarity assessment results and designates it as the second target fungal feature. This feature represents the standard feature that best matches the texture features formed by staining with characteristic metabolites in the sample to be tested within the second fungal feature database, thereby determining the survival status and related species information of the fungi in the sample.

[0088] For example, after evaluating the similarity of all storage layers in the second fungal feature database, it was found that the storage layer containing the texture feature of a certain metabolite in the live state of Candida albicans had the highest similarity score, 85 points (for example, the scores of other storage layers were all lower than this). Therefore, the fungal detection device determines that the second target fungal feature is the texture feature formed by staining with this specific metabolite of Candida albicans in the live state. This feature will be used to determine information such as the survival status of the fungus in the sample solution to be tested.

[0089] This invention can accurately screen from a second fungal feature database to identify the second target fungal feature that best matches the staining texture of characteristic metabolites in the sample to be tested. This provides a precise basis for determining the viability and species of fungi in the sample solution. Compared to random searching or methods without detailed feature extraction and stratified comparison, this improves the accuracy and efficiency of determining the viability and species of fungi.

[0090] In one embodiment, steps 501 to 503 are described as follows:

[0091] Step 501: Map the features of the first target fungus to the corresponding network nodes of the phylogenetic tree. The phylogenetic tree is a feature relation network constructed based on the first fungal feature database. The network nodes in the feature relation network represent different fungal species, and the edges represent the similarity associations of features between different species.

[0092] Optionally, the phylogenetic tree used by the fungal detection device is a feature relationship network constructed based on a first fungal feature database. The network nodes of this phylogenetic tree represent different fungal species, and the edges between nodes reflect the similarity and association of features among different fungal species. The device matches and maps the previously obtained first target fungal features with the nodes in the phylogenetic tree. By analyzing the various parameters of the first target fungal features, such as the specific values ​​or descriptions of morphology and color, the device finds the network node in the phylogenetic tree that best matches the target fungal feature; this node initially corresponds to a possible fungal species.

[0093] For example, the characteristics of the first target fungus, after previous matching, are determined to be highly similar to the standard characteristics of Candida albicans in the first fungal characteristic database. The fungal detection device searches for network nodes with similar Candida albicans characteristic descriptions in the phylogenetic tree constructed based on this database. Each node in the phylogenetic tree stores detailed characteristic information of the corresponding fungal species. By comparing features, the device finds that the fungal species represented by a specific node in the phylogenetic tree has morphological characteristics (elliptical cells within a certain size range) and color characteristics (hue, saturation, etc. under specific staining) that are almost identical to the characteristics of the first target fungus. Therefore, the characteristics of the first target fungus are mapped to this network node, which corresponds to the Candida albicans species.

[0094] Step 502: Based on the network nodes of the first target fungal feature, traverse along the edges of the evolutionary tree. By comparing the position of the first target fungal feature on the evolutionary tree with the distribution of known fungal species, and combining the topological structure and branching evolution rules of the evolutionary tree, determine the degree of matching between the first target fungal feature and the features of each node.

[0095] Furthermore, the fungal detection device starts from the network node of the mapped first target fungal feature and traverses along the edges of the evolutionary tree. During the traversal, it combines the topological structure of the evolutionary tree (such as the hierarchical relationship of nodes, the number and direction of branches, etc.) and the evolutionary rules of branches (such as the trend of feature changes from simple to complex, from primitive to evolved), comparing the position of the first target fungal feature on the evolutionary tree with the distribution of known fungal species. By analyzing the degree of difference between the features of different nodes and the first target fungal feature, the degree of matching between them is determined. For example, nodes that are closer to the mapped node and closely connected in evolutionary relationship may have a higher degree of matching with the first target fungal feature; while nodes that are farther away and on different branches may have a lower degree of matching.

[0096] Continuing with the example of mapping the features of the first target fungus to the node containing *Candida albicans*, the fungal detection device starts from this node and traverses along the edges of the phylogenetic tree towards adjacent nodes. In the phylogenetic tree, nodes adjacent to *Candida albicans* may represent other *Candida* species that are evolutionarily close to it. The device compares the features of these adjacent nodes with those of the first target fungus and finds that a certain adjacent node represents a fungal species that is morphologically similar to *Candida albicans*, but differs in some subtle structures (such as minor differences in cell wall thickness); in terms of color, the hue is slightly different. According to the topological structure of the phylogenetic tree, these adjacent nodes are on the same branch as the *Candida albicans* node and are relatively close. Through comprehensive analysis, the device determines that the features of these adjacent nodes match the features of the first target fungus to some extent, but not as well as the *Candida albicans* node itself. Continuing to traverse further nodes, it finds that fungal species on other branches have significantly different morphological and color features from the first target fungus, resulting in a lower degree of matching.

[0097] Step 503: Determine the types of fungi in the sample solution to be tested based on the degree of matching between the first target fungal features and the features of each node.

[0098] Furthermore, the fungal detection device determines the species of fungi in the sample solution based on the degree of matching between the characteristics of the first target fungus and the characteristics of each node. Among all the nodes traversed, the fungal species represented by the node with the highest degree of matching is the species of fungi in the sample solution. This is because the characteristics of this node are most similar to the characteristics of the first target fungus, which are obtained by matching with a large number of standard features in a first fungal feature database, thus possessing high reliability.

[0099] For example, after traversing and comparing in step 502, the fungal detection device found that the node containing *Candida albicans* had the highest degree of matching with the characteristics of the first target fungus, while the matching degree of other nodes was lower. Therefore, the device determined that the fungus in the sample solution to be tested was *Candida albicans*. This result, based on a comprehensive analysis and comparison of the characteristics of numerous nodes on the phylogenetic tree, ensures the accuracy of fungal species identification.

[0100] The embodiments of the present invention can accurately determine the species of fungi in the sample solution to be tested by means of the feature relationship network of the phylogenetic tree. Therefore, by using the structure and evolutionary rules of the phylogenetic tree, the relationship and differences between fungal characteristics can be analyzed more comprehensively and in-depth. This makes the determination of fungal species no longer limited to the simple matching of surface features, but a comprehensive consideration from the perspective of evolution, which improves the accuracy and scientific nature of fungal species judgment.

[0101] In one embodiment, steps 504 to 506 are described as follows:

[0102] Step 504: Determine the periodicity and repetition of the texture, as well as the arrangement pattern of texture units, based on the characteristics of the second target fungus.

[0103] Optionally, the fungal detection device, targeting the characteristics of the second target fungus, focuses on analyzing the periodicity and repetitiveness of the texture formed by staining with characteristic metabolites, as well as the arrangement pattern of texture units. Periodicity refers to whether the texture exhibits regular repetition within a certain area; repetitiveness is a quantitative description of periodicity, reflecting information such as the frequency of repetition. Texture units are the basic elements constituting the texture, and their arrangement patterns include ordered and disordered arrangements. By analyzing these texture characteristics, key information reflecting the metabolic state of the fungus can be obtained, because the production and distribution of metabolites differ between live and dead fungi, leading to differences in texture characteristics.

[0104] For example, the second target fungal feature corresponds to the texture features of Candida albicans metabolites. The fungal detection device analyzes this texture image. Observation reveals that within a certain image area, there exists a ring-like texture unit that repeats at regular intervals, exhibiting a clear periodicity. Further measurement shows that this ring-like texture unit reappears approximately every 5 pixels, confirming its repetitiveness. Simultaneously, the device analyzes the arrangement pattern of the ring-like texture units, finding that they are roughly arranged in an orderly manner according to a certain direction and interval, rather than being random. These are the analytical results regarding the texture's periodicity, repetition, and arrangement pattern.

[0105] Step 505: Based on the periodicity and repetition of texture, and the arrangement pattern of texture units, a matching is performed in the survival state feature knowledge base to obtain the state feature matching result. The survival state feature knowledge base is a knowledge base about the survival state characteristics of fungi, constructed with the second fungal feature database as its core and combined with relevant microbiological research results. The survival state feature knowledge base contains the typical manifestations and changing patterns of the texture features of metabolites of different fungi in both live and dead states.

[0106] Furthermore, the fungal detection device matches the periodicity, repetition, and arrangement patterns of texture units in a survival state feature knowledge base. This knowledge base, centered on a second fungal feature database, incorporates numerous relevant microbiological research findings, detailing the typical manifestations and changes in the texture features of metabolites from different fungi in both live and dead states. The device compares the input texture features with the stored texture feature standards for various fungi in different survival states to find the best match, thus obtaining the state feature matching result.

[0107] Continuing with the example of Candida albicans, the fungal detection device searches for matches in the viable state feature knowledge base based on the previously analyzed characteristics of Candida albicans texture, such as periodicity (repetition every 5 pixels), repetition, and ordered arrangement. The knowledge base describes the texture features of Candida albicans in its viable state as follows: specific texture units exist, exhibiting obvious periodicity, with repetition intervals between 4-6 pixels, and the arrangement is ordered. In the dead state, however, texture units typically become irregular, periodicity weakens or even disappears, and the arrangement is chaotic. Comparison reveals that the texture features of the Candida albicans to be detected are closer to the description in the knowledge base for its viable state.

[0108] Step 506: Determine the survival status of fungi in the sample solution to be tested based on the state feature matching results.

[0109] Furthermore, the fungal detection device determines the viability of the fungi in the sample solution based on the state feature matching results. If the state feature matching result is highly consistent with the feature description of a certain viability state (live or dead fungi) in the knowledge base, then the fungi in the sample solution can be determined to be in that viable state. This determination is based on the accurate analysis of the texture features and the precise matching with the standard features in the knowledge base. For example, since the state feature matching result in step 505 shows that the texture features of the Candida albicans to be detected are more consistent with the description of the live state, the fungal detection device determines that the Candida albicans in the sample solution is in a viable state.

[0110] The embodiments of the present invention can accurately determine the survival status of fungi in the sample solution to be tested. By using detailed analysis of texture features and matching with a professional knowledge base, the judgment results no longer rely solely on simple observation, but are based on scientific feature analysis and the support of a large number of research results, thereby improving the accuracy and reliability of judging the survival status of fungi.

[0111] The image recognition-assisted fungal detection device provided by the present invention will be described below. The image recognition-assisted fungal detection device described below can be referred to in correspondence with the image recognition-assisted fungal detection method described above. Figure 2 This is a schematic diagram of the image recognition-assisted fungal detection device provided by the present invention. The image recognition-assisted fungal detection device includes:

[0112] The image acquisition module 210 is used to acquire the image of the sample liquid to be tested based on the image acquisition device; the image to be tested is the image acquired after the sample liquid to be tested has been stained with fungal triple fluorescent staining solution;

[0113] The image fusion module 220 is used to fuse the images to be detected according to the acquisition angle to obtain the target fused image;

[0114] The feature extraction module 230 is used to extract features from the fused image to obtain feature extraction results; the feature extraction results include fungal morphological features, color features, and texture features formed by staining with characteristic metabolites;

[0115] The feature matching module 240 is used to match the first fungal feature database based on the morphological and color features of fungi to obtain the first target fungal feature, and to match the second fungal feature database based on the texture features formed by staining of characteristic metabolites to obtain the second target fungal feature. The first fungal feature database contains standard features of various fungi under specific culture and staining conditions, and the second fungal feature database contains texture features formed by staining of metabolites of live fungi and texture features formed by staining of metabolites of dead fungi.

[0116] The fungal detection module 250 is used to determine the type of fungus in the sample solution to be tested based on the characteristics of a first target fungus, and to determine the survival status of the fungus in the sample solution to be tested based on the characteristics of a second target fungus.

[0117] This invention, through multi-angle, high-resolution image acquisition, can obtain more comprehensive information about fungal samples, avoiding missed diagnoses due to low fungal quantity or atypical morphology. During image preprocessing, segmentation, and feature extraction, feature extraction comprehensively assesses the fungal morphological characteristics, color features, and texture features formed by staining with characteristic metabolites. This accurately highlights fungal features, unlike microscopic observation which struggles to distinguish between live and dead fungi. Therefore, compared to manual observation, feature extraction and matching based on objective image recognition algorithms avoids subjectivity and fatigue-induced misjudgments in manual image reading, greatly improving the accuracy of detection results and enabling accurate and rapid detection and identification of fungi.

[0118] Please see Figure 3 , Figure 3 This is a schematic diagram illustrating an embodiment of the electronic device provided in this invention. For example... Figure 3 As shown, this embodiment of the invention provides an electronic device 300, including a memory 310, a processor 320, and a computer program 311 stored in the memory 310 and executable on the processor 320. When the processor 320 executes the computer program 311, it performs the following steps:

[0119] The image to be tested is acquired using an image acquisition device; the image to be tested is obtained after the sample solution to be tested has been stained with a fungal triple fluorescent staining solution.

[0120] The images to be detected are fused according to the acquisition angle to obtain the target fused image;

[0121] Feature extraction is performed on the fused image to obtain the feature extraction results; the feature extraction results include fungal morphological features, color features, and texture features formed by staining by characteristic metabolites;

[0122] The first target fungal feature is obtained by matching the morphological and color features of fungi in the first fungal feature database, and the second target fungal feature is obtained by matching the texture features formed by staining of characteristic metabolites in the second fungal feature database. The first fungal feature database contains standard features of various fungi under specific culture and staining conditions, and the second fungal feature database contains texture features formed by staining of metabolites of live fungi and texture features formed by staining of metabolites of dead fungi.

[0123] The species of fungi in the sample solution to be tested are determined based on the characteristics of the first target fungus, and the survival status of the fungi in the sample solution to be tested is determined based on the characteristics of the second target fungus.

[0124] Please see Figure 4 , Figure 4 This is a schematic diagram illustrating an embodiment of a computer-readable storage medium provided in accordance with the present invention. For example... Figure 4 As shown, this embodiment provides a computer-readable storage medium 400 on which a computer program 311 is stored. When the computer program 311 is executed by a processor, it performs the following steps:

[0125] The image to be tested is acquired using an image acquisition device; the image to be tested is obtained after the sample solution to be tested has been stained with a fungal triple fluorescent staining solution.

[0126] The images to be detected are fused according to the acquisition angle to obtain the target fused image;

[0127] Feature extraction is performed on the fused image to obtain the feature extraction results; the feature extraction results include fungal morphological features, color features, and texture features formed by staining by characteristic metabolites;

[0128] The first target fungal feature is obtained by matching the morphological and color features of fungi in the first fungal feature database, and the second target fungal feature is obtained by matching the texture features formed by staining of characteristic metabolites in the second fungal feature database. The first fungal feature database contains standard features of various fungi under specific culture and staining conditions, and the second fungal feature database contains texture features formed by staining of metabolites of live fungi and texture features formed by staining of metabolites of dead fungi.

[0129] The species of fungi in the sample solution to be tested are determined based on the characteristics of the first target fungus, and the survival status of the fungi in the sample solution to be tested is determined based on the characteristics of the second target fungus.

[0130] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute the image recognition-assisted fungal detection method provided by the above methods. The image recognition-assisted fungal detection method includes:

[0131] The image to be tested is acquired using an image acquisition device; the image to be tested is obtained after the sample solution to be tested has been stained with a fungal triple fluorescent staining solution.

[0132] The images to be detected are fused according to the acquisition angle to obtain the target fused image;

[0133] Feature extraction is performed on the fused image to obtain the feature extraction results; the feature extraction results include fungal morphological features, color features, and texture features formed by staining by characteristic metabolites;

[0134] The first target fungal feature is obtained by matching the morphological and color features of fungi in the first fungal feature database, and the second target fungal feature is obtained by matching the texture features formed by staining of characteristic metabolites in the second fungal feature database. The first fungal feature database contains standard features of various fungi under specific culture and staining conditions, and the second fungal feature database contains texture features formed by staining of metabolites of live fungi and texture features formed by staining of metabolites of dead fungi.

[0135] The species of fungi in the sample solution to be tested are determined based on the characteristics of the first target fungus, and the survival status of the fungi in the sample solution to be tested is determined based on the characteristics of the second target fungus.

[0136] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0137] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0138] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims and equivalents of this invention, this invention also intends to include these modifications and variations.

Claims

1. A fungal detection method based on image recognition assistance, characterized in that, include: The image to be tested is acquired using an image acquisition device; The image to be detected is an image obtained after the sample solution to be detected has been stained with fungal triple fluorescent staining solution; The images to be detected are fused according to the acquisition angle to obtain the target fused image; Feature extraction is performed on the target fused image to obtain feature extraction results; the feature extraction results include fungal morphological features, color features, and texture features formed by staining with characteristic metabolites; The first target fungal feature is obtained by matching the morphological and color features of fungi in the first fungal feature database, and the second target fungal feature is obtained by matching the texture features formed by staining of characteristic metabolites in the second fungal feature database. The first fungal feature database contains standard features of various fungi under specific culture and staining conditions, and the second fungal feature database contains texture features formed by staining of metabolites of live fungi and texture features formed by staining of metabolites of dead fungi. The species of fungi in the sample solution to be tested are determined based on the characteristics of the first target fungus, and the survival status of the fungi in the sample solution to be tested is determined based on the characteristics of the second target fungus. The determination of the viability status of fungi in the sample solution to be tested based on the characteristics of the second target fungus includes: The periodicity and repetition of the texture, as well as the arrangement pattern of the texture units, are determined based on the characteristics of the second target fungus. Based on the periodicity and repetition of textures, as well as the arrangement pattern of texture units, a matching is performed in the survival state feature knowledge base to obtain the state feature matching result; the survival state feature knowledge base is a knowledge base about the survival state characteristics of fungi constructed by combining the second fungal feature database with relevant microbiological research results; the survival state feature knowledge base contains the typical manifestations and change patterns of the texture features of metabolites of different fungi in the live and dead states. The survival status of fungi in the sample solution to be tested is determined based on the state feature matching results.

2. The image recognition-assisted fungal detection method according to claim 1, characterized in that, The step of determining the type of fungus in the sample solution to be tested based on the characteristics of the first target fungus includes: The first target fungal features are mapped to the corresponding network nodes of the phylogenetic tree; the phylogenetic tree is a feature relationship network constructed based on the first fungal feature database; the network nodes in the feature relationship network are different fungal species, and the edges are similarity associations of features between different species; The network nodes based on the first target fungal feature traverse along the edges of the evolutionary tree. By comparing the position of the first target fungal feature on the evolutionary tree with the distribution of known fungal species, and combining the topological structure and branching evolution rules of the evolutionary tree, the degree of matching between the first target fungal feature and the features of each node is determined. The species of fungi in the sample solution to be tested are determined based on the degree of matching between the characteristics of the first target fungus and the characteristics of each node.

3. The image recognition-assisted fungal detection method according to claim 1, characterized in that, The first target fungal feature is obtained by matching fungal morphological and color features in a first fungal feature database, including: The fungal morphological features and color features are decomposed according to different dimensions to obtain the decomposed morphological and color feature components; Based on the decomposed morphological features, a preliminary comparison is made with the standard fungal morphological features of the same classification level in the first fungal feature database according to the type and combination of basic geometric elements to obtain morphological feature matching results; the classification level includes phylum, class, order, family, genus and species; Based on the color feature components, the color feature matching result is compared with the standard fungal color feature corresponding to the index position of the morphological feature matching result in the first fungal feature database to obtain the color feature matching result. The characteristics of the first target fungus are determined based on the color feature matching results.

4. The image recognition-assisted fungal detection method according to claim 1, characterized in that, The texture features formed based on staining with characteristic metabolites are matched with the second fungal feature database to obtain the features of the second target fungus, including: The primary and secondary directional information and structural information of the texture features formed by the staining of the characteristic metabolites are extracted based on the image structure tensor algorithm. The hierarchical storage structure of the second fungal feature database is traversed to determine the relevant information in each storage layer. The second fungal feature database is stored separately according to the texture features of live and dead fungal metabolites, and within each storage area, it is layered according to different fungal species and metabolite types. The similarity evaluation results are based on the similarity measurement algorithm to determine the primary and secondary directional information and structural information in the texture features and the relevant information in each storage layer. The characteristics of the second target fungus are determined based on the similarity assessment results.

5. The image recognition-assisted fungal detection method according to any one of claims 1 to 4, characterized in that, The step of fusing the images to be detected according to the acquisition angle to obtain the target fused image includes: Based on the acquisition angle, each image to be detected is grouped according to a preset angle range to obtain multiple initial detection image groups; The images to be detected in each initial detection image group are arranged according to the size of the acquisition angle to obtain the updated detection image group; For each image in the updated detection image group, key feature points of each image to be detected are extracted based on the feature extraction algorithm; key feature points represent important structural and texture information of the image; The target fused image is obtained by fusing the key feature points of each image to be detected in each updated detection image group.

6. The image recognition-assisted fungal detection method according to claim 5, characterized in that, The process of fusing key feature points of each image to be detected in each updated detection image group to obtain the target fused image includes: For each image in the updated detection image group, the correspondence between different images is determined based on the position and descriptor of the key feature points of each image to be detected, and the image matching result is obtained; the image matching result indicates that the feature points represent the same object or region in different images. Based on the image matching results, the overlapping regions between images within each updated group of detected images are determined; the overlapping regions represent the parts that are commonly covered in different images. For images within the same updated detection image group, an image fusion algorithm is used in the overlapping area to fuse the images to be detected, generating a fused image within the group; For images between different updated detection image groups, the images within each updated detection image group are fused again to obtain the target fused image.

7. A fungal detection device based on image recognition assistance, characterized in that, include: The image acquisition module is used to acquire images of the sample liquid to be tested based on the image acquisition device; The image to be detected is an image obtained after the sample solution to be detected has been stained with fungal triple fluorescent staining solution; The image fusion module is used to fuse the images to be detected according to the acquisition angle to obtain the target fused image; The feature extraction module is used to extract features from the target fused image to obtain feature extraction results; the feature extraction results include fungal morphological features, color features, and texture features formed by staining with characteristic metabolites; The feature matching module is used to match the first fungal feature database based on the morphological and color features of fungi to obtain the first target fungal feature, and to match the second fungal feature database based on the texture features formed by staining of characteristic metabolites to obtain the second target fungal feature. The first fungal feature database contains standard features of various fungi under specific culture and staining conditions, and the second fungal feature database contains texture features formed by staining of metabolites of live fungi and texture features formed by staining of metabolites of dead fungi. The fungal detection module is used to determine the species of fungi in the sample solution to be tested based on the characteristics of the first target fungus, and to determine the survival status of the fungi in the sample solution to be tested based on the characteristics of the second target fungus. The determination of the viability status of fungi in the sample solution to be tested based on the characteristics of the second target fungus includes: The periodicity and repetition of the texture, as well as the arrangement pattern of the texture units, are determined based on the characteristics of the second target fungus. Based on the periodicity and repetition of textures, as well as the arrangement pattern of texture units, a matching is performed in the survival state feature knowledge base to obtain the state feature matching result; the survival state feature knowledge base is a knowledge base about the survival state characteristics of fungi constructed by combining the second fungal feature database with relevant microbiological research results; the survival state feature knowledge base contains the typical manifestations and change patterns of the texture features of metabolites of different fungi in the live and dead states. The survival status of fungi in the sample solution to be tested is determined based on the state feature matching results.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the image recognition-assisted fungal detection method as described in any one of claims 1-6.

9. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the image recognition-assisted fungal detection method as described in any one of claims 1-6.

Images