Computer Vision-Based Longhorn Beetle Monitoring Method and System for Ancient and Famous Trees

Abstract

The present invention provides a monitoring method and system for long-lived famous trees' longhorn beetles based on computer vision, which relates to the field of computer vision technology. By collecting ultraviolet fluorescence images under multiple exposure modes and combining deep learning technology for semantic segmentation and fluorescence feature extraction, the present invention establishes an association model between fluorescence features and the number of longhorn beetles. This technical solution can effectively improve the monitoring ability in low-light environments and provide more accurate and quantitative predicted values for the number of longhorn beetles. At the same time, through the trend analysis of fluorescence features in consecutive time periods, an adjustment strategy for the exposure mode of subsequent ultraviolet fluorescence images is provided.

Description

Technical Field

[0001] The present invention relates to the field of computer vision technology, and specifically to a monitoring method and system for longhorn beetles on ancient and famous trees based on computer vision. Background Art

[0002] In the process of protecting ancient and famous trees, the harm of longhorn beetles is particularly prominent. As a major boring pest in the forestry ecosystem, the boring behavior of the larvae of longhorn beetles not only directly damages the xylem structure of the trees, resulting in damage to the mechanical support and conduction systems, but also creates an infection channel for secondary pathogenic microorganisms, thereby triggering systemic diseases and ultimately possibly leading to the death of the host plant. Therefore, timely and accurately monitoring the activities of longhorn beetles can provide an important basis for the management and protection of ancient and famous trees. In recent years, plant pest and disease monitoring technologies based on computer vision have gradually emerged, and through means such as image processing and deep learning, accurate identification and monitoring of plant pests and diseases have been achieved.

[0003] In the prior art, the publication number is CN117576728A, and the name is a monitoring method and system for Monochamus alternatus based on computer vision. The method includes obtaining a plurality of first images of a target area; using the YOLO-v8 object detection algorithm to perform a first identification of Monochamus alternatus in each first image, and marking the first identification result with a detection frame; inputting the second image in the detection frame into an adversarial convolutional autoencoder to extract the deep features of the second image; inputting the deep features of the second image into a Gaussian process classifier to perform a second identification of Monochamus alternatus in the second image, and obtaining the number of Monochamus alternatus in the first image. Through image acquisition and image recognition, the identification and quantity statistics of Monochamus alternatus are realized, providing effective pest control support for forestry management.

[0004] Existing monitoring technologies mostly rely on visible light imaging and are difficult to effectively extract the weak signals of longhorn beetle activities. Especially at night or in low-light environments, traditional imaging technologies are easily affected by insufficient light, resulting in missing and inaccurate monitoring data. In addition, existing technologies usually lack effective temporal analysis and do not fully utilize the information in the time dimension to capture changes in the number and activity patterns of longhorn beetles. Even in some cases, fluorescence imaging means are adopted, and there are also deficiencies in the processing and analysis of fluorescence signals, often only providing qualitative observations, lacking quantitative analysis and model establishment.

[0005] The above information disclosed in the above background art section is only used to strengthen the understanding of the background of the present disclosure, and therefore it may include information that does not constitute the prior art known to those of ordinary skill in the art. Summary of the Invention

[0006] The object of the present invention is to provide a monitoring method and system for long-lived famous trees and ancient trees with longicorn beetles based on computer vision, so as to solve the problems raised in the above-mentioned background technology.

[0007] To achieve the above object, the present invention provides the following technical solutions:

[0008] A monitoring method for long-lived famous trees and ancient trees with longicorn beetles based on computer vision, the specific steps include:

[0009] Step S1: In the night or low-light environment, initially use the medium exposure mode within a variety of exposure modes to continuously collect the ultraviolet fluorescence images generated by the activities of longicorn beetles on the surface of the current long-lived famous trees and ancient trees, and record the collected ultraviolet fluorescence images as sequential image data;

[0010] Step S2: Preprocess the ultraviolet fluorescence image data in the sequential image data;

[0011] Step S3: Use a deep learning model to perform semantic segmentation on the preprocessed ultraviolet fluorescence image, and extract the fluorescence regions related to the excrement and activity traces of longicorn beetles;

[0012] Step S4: Obtain the segmented fluorescence regions and perform extraction and analysis of fluorescence features. The fluorescence features include the area of the fluorescence region and the average fluorescence signal intensity;

[0013] Step S5: Obtain the historical data of the number of longicorn beetles under different fluorescence features, and establish an association model between the fluorescence features of the ultraviolet fluorescence image and the number of longicorn beetles according to the fluorescence features of the fluorescence region;

[0014] Extract the fluorescence features of the latest ultraviolet fluorescence image in the sequential image data; input the extracted fluorescence features into the established association model, and finally obtain the predicted value of the number of longicorn beetles based on the medium exposure mode;

[0015] Step S6: Perform trend analysis of the fluorescence features of the currently collected sequential image data for a continuous time period to generate a fluorescence feature change index, and perform combined analysis on the fluorescence feature change index and the predicted value of the number of longicorn beetles. The analysis results are used to make corresponding adjustments to the exposure mode of the subsequent ultraviolet fluorescence images.

[0016] A monitoring system for long-lived famous trees and ancient trees with longicorn beetles based on computer vision, the system is used to execute the monitoring method for long-lived famous trees and ancient trees with longicorn beetles based on computer vision, including:

[0017] Sequential image acquisition module: used to continuously collect the ultraviolet fluorescence images generated by the activities of longicorn beetles on the surface of the current long-lived famous trees and ancient trees in the night or low-light environment, initially using the medium exposure mode within a variety of exposure modes, and record the collected ultraviolet fluorescence images as sequential image data;

[0018] Preprocessing module: used to preprocess the ultraviolet fluorescence image data in the time-series image data;

[0019] Fluorescent region determination module: used to perform semantic segmentation on the preprocessed ultraviolet fluorescence image by using a deep learning model, and extract the fluorescent regions related to the excrement and activity traces of longhorn beetles;

[0020] Extraction module: used to obtain the segmented fluorescent regions and perform extraction and analysis of fluorescent features, where the fluorescent features include the area of the fluorescent region and the average fluorescent signal intensity;

[0021] Association model construction module: used to obtain the historical data of the number of longhorn beetles under different fluorescent features, and establish an association model between the fluorescent features of the ultraviolet fluorescence image and the number of longhorn beetles according to the fluorescent features of the fluorescent region;

[0022] Extract the fluorescent features of the latest ultraviolet fluorescence image in the time-series image data; input the extracted fluorescent features into the established association model, and finally obtain the predicted value of the number of longhorn beetles based on the medium exposure mode;

[0023] Judgment and adjustment module: used to perform trend analysis of the fluorescent features of the time-series image data collected currently for a continuous time period to generate a fluorescent feature change index, and perform combined analysis on the fluorescent feature change index and the predicted value of the number of longhorn beetles, and the analysis result is used to make corresponding adjustments to the exposure mode of the subsequent ultraviolet fluorescence image.

[0024] Compared with the prior art, the beneficial effects of the present invention are: by collecting ultraviolet fluorescence images under multiple exposure modes, combining deep learning technology for semantic segmentation and fluorescent feature extraction, an association model between the fluorescent features and the number of longhorn beetles is established; this technical solution can effectively improve the monitoring ability in low-light environments and provide more accurate and quantitative predicted values of the number of longhorn beetles; at the same time, through trend analysis of the fluorescent features for a continuous time period, this method helps to identify the dynamic changes of longhorn beetle activities, and further provides an adjustment strategy for the exposure mode of the subsequent ultraviolet fluorescence image. Brief description of the drawings

[0025] Figure 1 It is a schematic diagram of the overall method flow of the present invention;

[0026] Figure 2 It is a block diagram of the overall system module of the present invention. Detailed implementation manners

[0027] To make the objectives, technical solutions and advantages of the present invention clearer and more understandable, the following further details the present invention with reference to specific embodiments.

[0028] It should be noted that, unless otherwise defined, the technical terms or scientific terms used in the present invention should have the ordinary meanings understood by those with ordinary skills in the field to which the present invention belongs. The "first", "second" and similar words used in the present invention do not denote any order, quantity or importance, but are only used to distinguish different components. Words such as "including" or "comprising" mean that the elements or objects appearing before this word cover the elements or objects listed after this word and their equivalents, without excluding other elements or objects. Words such as "connected" or "linked" are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. "Up", "down", "left", "right", etc. are only used to represent relative positional relationships. When the absolute position of the object being described changes, the relative positional relationship may also change accordingly.

[0029] Embodiment 1:

[0030] Please refer to Figure 1 , the present invention provides a technical solution:

[0031] A monitoring method for longicorn beetles on ancient and famous trees based on computer vision, the specific steps include:

[0032] Step S1: At night or in low-light environments, initially use the medium exposure mode within multiple exposure modes to continuously collect ultraviolet fluorescence images generated by the activities of longicorn beetles on the surface of the current ancient and famous trees, and record the collected ultraviolet fluorescence images as sequential image data;

[0033] Further explanation: The collection of the sequential image data includes:

[0034] Set the night as the time period represented by 19:00 to 6:00 the next day, and set the low-light environment as the condition where the light intensity is less than 5 lux;

[0035] Use an ultraviolet fluorescence excitation imaging system to arrange multiple ultraviolet fluorescence cameras around the ancient and famous trees, and then conduct continuous image sampling on the surface of the ancient and famous trees at different time periods to obtain complete ultraviolet fluorescence images of the surface of the ancient and famous trees;

[0036] The ultraviolet fluorescence excitation imaging system includes an ultraviolet fluorescence camera and an excitation light source. The excitation light source uses an LED ultraviolet lamp with a wavelength in the range of 365 nm to 395 nm. In the wavelength range of 365 nm to 395 nm, the excrement of Monochamus alternatus can produce a fluorescence effect;

[0037] The ultraviolet fluorescence camera uses an industrial-grade ultraviolet fluorescence camera of the FLIR UV series. The acquisition angle of the ultraviolet camera forms a 45° angle with the target surface to be photographed. Each ultraviolet fluorescence camera is 1.5 to 2 meters away from the target to be photographed, so as to reduce light interference and improve the response intensity of the fluorescence signal. In this embodiment, 4 ultraviolet fluorescence cameras are set;

[0038] The multiple exposure modes include: low exposure, medium exposure, and high exposure;

[0039] Set the exposure time range of the multiple exposure modes within the interval [BG1, BG2], where BG1 and BG2 are the minimum exposure time and the maximum exposure time respectively, and the exposure times of low exposure, medium exposure, and high exposure increase in sequence;

[0040] Set i ∈ {1, 2, …, m}, where i represents the index of the acquisition times, m represents the total number of acquisitions of the ultraviolet fluorescence images in the time-series image data, j ∈ {1, 2, 3}, where j represents the type index of the exposure mode, and j = 1 represents the low exposure mode; j = 2 represents the medium exposure mode; j = 3 represents the high exposure mode. In each acquisition, it is initially set to perform the acquisition of the ultraviolet fluorescence image in the medium exposure mode;

[0041] In the time-series image data, the ultraviolet fluorescence image in the j-th exposure mode of the i-th acquisition is denoted as i(j);

[0042] Divide the acquisition times in {1, 2, …, m} into at least two parts. The first part represents the acquisition times in the first time period, and the second part represents the acquisition times in the second time period, and the first time period and the second time period form a continuous time period;

[0043] In this embodiment, the example values of BG1 and BG2 are taken as 10 milliseconds and 100 milliseconds respectively, where:

[0044] Low exposure is an exposure time less than 10 milliseconds;

[0045] Medium exposure is an exposure time in the interval [10, 100] milliseconds;

[0046] High exposure is an exposure time exceeding 100 milliseconds;

[0047] Each ultraviolet fluorescence camera works continuously at a frequency of 10 minutes per frame for 11 hours to collect spectral imaging data covering the entire area of ancient and famous trees; 11 hours is adjusted according to the condition of "light intensity less than 5 lux";

[0048] The above acquisition process is selected to be measured during a period with clear weather, moderate humidity, and no precipitation;

[0049] For the determination of moderate humidity: Monitor the humidity in the places where longhorn beetles are active, collect long-term humidity data, and analyze the most suitable humidity range in combination with the data on the activity traces of longhorn beetles.

[0050] Laboratory experiments can simulate different humidity conditions to evaluate the activity intensity of longhorn beetles and their fluorescence signal realization;

[0051] In a controllable environment such as a greenhouse, artificially adjust the humidity (between 30% and 70%) to conduct experiments on the activity of longhorn beetles and the simulation of fluorescence signals.

[0052] It should be noted that:

[0053] The activities of longhorn beetles, including gnawing on tree bark, excreting, secreting saliva, and other behaviors, will generate or enhance the fluorescence signal on the tree surface;

[0054] The excreta and saliva of longhorn beetles contain fluorescent active substances that can emit fluorescence under ultraviolet light irradiation;

[0055] Step S2: Preprocess the ultraviolet fluorescence image data in the time-series image data;

[0056] Further explanation: The preprocessing of the ultraviolet fluorescence image data specifically includes:

[0057] Use an image registration algorithm to register and splice the images of ancient and famous trees taken by multiple ultraviolet fluorescence cameras from different perspectives to generate a complete ultraviolet fluorescence image of the surface of the ancient and famous trees;

[0058] Based on the bilateral filtering algorithm, remove the low-frequency noise in the ultraviolet fluorescence image and enhance the high-frequency edge features at the same time;

[0059] Perform histogram equalization on the denoised ultraviolet fluorescence image to enhance the contrast between the fluorescence region and the background.

[0060] For the implementation steps of image registration and splicing, as well as denoising and image enhancement, they are as follows:

[0061] 1.1) Feature point extraction:

[0062] Open each ultraviolet fluorescence image, use image processing software (such as ImageJ, Photoshop, or MATLAB) or supporting tools to run the SIFT algorithm (the OpenCV library can be used), and extract key point features on each ultraviolet fluorescence image through the SIFT algorithm; The determination process of the key point features is as follows:

[0063] Create a scale space through Gaussian blur: Perform Gaussian blur processing on the ultraviolet fluorescence images of different scales to generate a series of scale space images.

[0064] Detecting extreme points: In the generated scale-space image, detect local extreme points; the extreme points are potential key points.

[0065] Feature descriptor generation:

[0066] Direction assignment: Assign one or more directions to each extreme point to ensure the descriptor has rotational invariance.

[0067] Calculating the descriptor: Generate a rotation-invariant 128-dimensional vector descriptor based on the local neighborhood gradient of the extreme point.

[0068] Use MATLAB, the OpenCV library in Python, or other image processing software to actually execute the above steps.

[0069] 1.2) Feature point matching:

[0070] In the software tool, select the medium-exposure ultraviolet fluorescence image as the reference image; perform feature point matching between all ultraviolet fluorescence images and the reference image.

[0071] Use the RANSAC algorithm to filter out incorrect matching points to ensure accurate matching.

[0072] 1.3) Geometric transformation:

[0073] Calculate the transformation matrix for each ultraviolet fluorescence image. This embodiment includes translation, rotation, and scaling parameters.

[0074] Apply this transformation matrix in the software to align the images to the reference image.

[0075] Ensure that the feature positions coincide between images from different perspectives.

[0076] 1.4) Image stitching:

[0077] In the graphics editing software, gradually stitch the aligned images into a complete ultraviolet fluorescence image, using a blending method to smooth the image edge transition.

[0078] 2) For noise removal:

[0079] 2.1) Parameter selection:

[0080] Use the bilateral filtering function supported by the image editing software to set the filtering parameters.

[0081] Set the spatial domain parameter σd to 10 and the intensity domain parameter σr to 0.1. These settings are directly set in professional image processing tools.

[0082] 2.2) Applying the filter:

[0083] Perform bilateral filtering, scan the entire image, including complex and detailed areas, weaken background noise while preserving boundary sharpness.

[0084] 3) For image enhancement:

[0085] 3.1) Histogram calculation:

[0086] Use the histogram function built into the graphics editing software to generate the grayscale histogram of the denoised image.

[0087] 3.2) Equalization processing:

[0088] Execute the histogram equalization function in the software.

[0089] Adjust the contrast of the ultraviolet fluorescence image to improve the relative visibility of the fluorescence area to the background.

[0090] Save the processed image in a high-quality format (JPEG / PNG), maintaining the original resolution.

[0091] Step S3: Use a deep learning model to perform semantic segmentation on the preprocessed ultraviolet fluorescence image, and extract the fluorescence areas related to longhorn beetle excrement and activity traces;

[0092] Further explanation: The deep learning model completes pixel-level image analysis through the semantic segmentation model UNet, where the pre-trained model parameters are used to classify the fluorescence areas;

[0093] Label the fluorescence areas in the segmented results in the form of a classification mask to distinguish the fluorescence areas from the non-fluorescence areas, where the fluorescence areas include the longhorn beetle excrement area and the longhorn beetle activity trace area;

[0094] Generate a multi-channel two-dimensional segmentation mask for the fluorescence area for subsequent fluorescence feature extraction.

[0095] The acquisition method of the pre-trained model parameters is as follows:

[0096] Model training data preparation:

[0097] The dataset contains ultraviolet fluorescence images with manual annotations, and the annotation content includes:

[0098] Fluorescence area: Longhorn beetle excrement and activity trace features;

[0099] Non-fluorescence area: Invalid information such as bark and background.

[0100] Collect no less than 5000 ultraviolet fluorescence image sample data with manual annotations, and divide them into a training set (80%) and a validation set (20%) to ensure the generalization ability of the model.

[0101] Normalize all images to ensure consistent dimensions and color spaces of the input images.

[0102] Select UNet as the base model for semantic segmentation;

[0103] Initialization of pre-trained parameters:

[0104] Using transfer learning techniques, select ResNet as the backbone network of the encoder and load its pre-trained parameters to accelerate the convergence speed.

[0105] Effectively combine the advanced feature representation ability of ResNet through transfer learning to improve the accuracy of fluorescence region recognition.

[0106] Training process:

[0107] Use the cross-entropy loss function combined with the dice coefficient to optimize the objective, clearly distinguishing the high-signal fluorescence region from the background region.

[0108] Model hyperparameter settings: The learning rate is 1e-4, the batch size is 16, and it is initialized with ResNet pre-trained parameters through transfer learning techniques.

[0109] Use Stochastic Gradient Descent (SGD) or Adam optimizer in the training set to adjust the parameters.

[0110] Regularly record the training and validation errors to prevent overfitting, and perform data augmentation (such as rotation, scaling) as needed;

[0111] Analysis of segmentation results:

[0112] Use the trained UNet model to perform semantic segmentation on the validation set and new images.

[0113] Each segmented region is represented by a different color: red represents the excrement of longhorn beetles, and blue represents the activity traces, facilitating intuitive analysis.

[0114] Generate a multi-channel two-dimensional segmentation mask:

[0115] Convert the segmentation result into a multi-channel two-dimensional mask, with each channel corresponding to a classification (for example, excrement of longhorn beetles and activity traces).

[0116] These masks are used for subsequent fluorescence signal feature extraction to improve the analysis accuracy of different biological features.

[0117] Step S4: Obtain the segmented fluorescence regions and perform extraction and analysis of fluorescence features. The fluorescence features include the area of the fluorescence region and the average fluorescence signal intensity;

[0118] Further explanation: Define the area of the fluorescence region of the ultraviolet fluorescence image i(j) as Af,i(j) ; Calculate A by counting the total number of pixels in the segmentation mask f,i(j) ;

[0119] Based on the fluorescence area, define the average fluorescence signal intensity of the fluorescence area as F int,i(j) , and the formula is:

[0120]

[0121] where P(x, y) represents the gray value of the pixel at pixel coordinates (x, y);

[0122] (x, y) ∈ A f,i(j) means that the pixel coordinates (x, y) belong to the set of all pixel point coordinates within the fluorescence area A f,i(j) within.

[0123] Step S5: Obtain the historical data of the number of longhorn beetles under different fluorescence characteristics, and establish an association model between the fluorescence characteristics of the ultraviolet fluorescence image and the number of longhorn beetles according to the fluorescence characteristics of the fluorescence area;

[0124] Extract the fluorescence characteristics of the latest ultraviolet fluorescence image in the time-series image data; input the extracted fluorescence characteristics into the established association model, and finally obtain the predicted value of the number of longhorn beetles based on the medium exposure mode;

[0125] In the night or low-light environment, collect an ultraviolet fluorescence image dataset of ancient and famous trees covering the active season of longhorn beetles; each ultraviolet fluorescence image records the acquisition time, the i-th acquisition, and the j-th exposure mode; it also includes the record of the number of longhorn beetles corresponding to the manually marked fluorescence feature data;

[0126] Extract the fluorescence characteristics of each ultraviolet fluorescence image in the fluorescence area; specifically, the average fluorescence signal intensity F int,i(j) and the fluorescence area A f,i(j) ;

[0127] Match and label the fluorescence feature data of the j-th exposure mode collected each time with the historically manually recorded number of longhorn beetles to generate a labeled dataset;

[0128] Verify the correspondence between the fluorescence characteristics extracted from the ultraviolet fluorescence image and the actually measured number of longhorn beetles by collecting calibration data through regular experiments to ensure the accuracy and reliability of the data.

[0129] Clean all the collected fluorescence feature data to remove outliers and noise data;

[0130] Standardize the data to ensure that the dimensions of the feature data are consistent, and adopt the z-score standardization method (i.e., the mean is 0 and the standard deviation is 1).

[0131] Based on the labeled dataset, a non-linear regression model is selected. The non-linear regression model is Support Vector Regression (SVR) or Random Forest Regression (RF), which is used to establish an association model between fluorescence features and the number of longhorn beetles.

[0132] Selection criteria: Support Vector Regression (SVR) is suitable for small sample datasets and has good generalization ability; Random Forest Regression (RF) is suitable for processing high-dimensional feature data and has high robustness.

[0133] Define the association model under the j-th exposure mode as:

[0134] N1 j =h j (F int，i(j) ,A fi(j) )

[0135] where h j is the non-linear function under the j-th exposure mode, which is used to describe the association characteristics between the fluorescence features and the number of longhorn beetles under the j-th exposure mode; N1 j is the predicted value of the number of longhorn beetles corresponding to the fluorescence features under the j-th exposure mode;

[0136] Extract the fluorescence features of the latest ultraviolet fluorescence image in the time-series image data. Denote the average fluorescence signal intensity and the fluorescence area included in the fluorescence features extracted under the j-th exposure mode in the n-th acquisition as F int,n(j) and A f,n(j) ; n represents the acquisition times index of the latest ultraviolet fluorescence image;

[0137] Use 50% of the collected data as the training set for training the association model, and use the cross-validation technique to adjust the model hyperparameters to optimize the model performance;

[0138] Use the remaining 50% of the collected data as the test set to evaluate the prediction accuracy and generalization ability of the model;

[0139] Input the corresponding F int,n(j) and A f,n(j) of the extracted fluorescence feature data into the trained association model, calculate and output the predicted value N1 of the number of longhorn beetles under the j-th exposure mode in the n-th acquisition n(j) ; Denote the predicted value of the number of longhorn beetles under the medium exposure mode as N1 n(2) .

[0140] Step S6: Conduct a fluorescence feature trend analysis on the time-series image data collected currently for a continuous time period to generate a fluorescence feature change index, and conduct a combined analysis on the fluorescence feature change index and the predicted value of the number of longhorn beetles. The analysis results are used to make corresponding adjustments to the exposure mode of subsequent ultraviolet fluorescence images;

[0141] Adjustment of the exposure mode is used to reduce the data storage burden of the ultraviolet fluorescence camera;

[0142] Further explanation: Denote the first time period and the second time period as T1 and T2 respectively, and represent T1 and T2 uniformly as T ∈ {T1, T2}; Differentially calculate the fluorescence characteristics within T1 and T2 to obtain the average fluorescence signal intensity and the average fluorescence area within the following first time period and second time period respectively:

[0143]

[0144] Among them, and represent the average fluorescence signal intensity under the first time period and the second time period respectively;

[0145] and represent the average fluorescence area under the first time period and the second time period respectively;

[0146] n T1 represents the number of acquisitions corresponding to the first time period; n - n T1 represents the number of acquisitions corresponding to the second time period; i = n T1 + 1 represents the characterization of the first acquisition number within the second time period;

[0147] Calculate the difference between the average fluorescence signal intensity of the second time period and the first time period to obtain the first difference value ΔF int,T ;

[0148] Calculate the difference between the average fluorescence area of the second time period and the first time period to obtain the second difference value ΔA f , T;

[0149] ΔF int,T and ΔA f , T calculation formulas are as follows:

[0150]

[0151] Combined with ΔF int,T and ΔA f , T, define the fluorescence characteristic change index as SI T , and the calculation formula is as follows:

[0152]

[0153] Wherein: α and β are weight coefficients, α + β = 1, and the value ranges of α and β are within the interval (0, 1); α and β are used to adjust the contributions of corresponding parameters to the fluorescence feature change index; in this embodiment, α = 0.6 and β = 0.4 are selected; the weights of α and β are determined by the entropy weight method and the fuzzy analytic hierarchy process (FAHP).

[0154] In this embodiment, 0.23 ≤ α ≤ 0.68; 0.11 ≤ β ≤ 0.59;

[0155] For the first difference value ΔF int,T :

[0156] ΔF int,T represents the change in the average fluorescence signal intensity within two time periods; if ΔF int,T increases, with the weight α remaining unchanged, SI T increases accordingly.

[0157] For the second difference value ΔA f ,T:

[0158] ΔA f ,T represents the change in the average fluorescence area within two time periods; if ΔA f ,T increases, with the weight β remaining unchanged, SI T also increases accordingly;

[0159] For the interactive adjustment of the weight coefficients α and β:

[0160] The sum of the weights of α and β is 1:

[0161] When α > β: it indicates that the change in the average fluorescence signal intensity contributes more to SI T .

[0162] When β > α: it indicates that the change in the average fluorescence area contributes more to SI T .

[0163] By adjusting the magnitudes of α and β, SI T can be flexibly customized according to the importance of parameters in specific scenarios. For example, when the activity of longhorn beetles has a greater impact on the fluorescence signal intensity, increase the value of α.

[0164] The default values recommended in this embodiment are α = 0.6 and β = 0.4, that is, the influence ratio of the change in fluorescence intensity on the index is 60%, and the influence of the change in area is 40%;

[0165] When ΔF int,T or ΔA f ,T increases:

[0166] If ΔF int,T and ΔAf , an increase in both will lead to a decrease in T , thus increasing the value of SI.

[0167] ΔF int,T and ΔA f respectively measure the differential changes in signal intensity and regional area. α and β adjust the relative importance of these two parameters through weight assignment. Real-time monitoring of ΔF int,T and ΔA f , and adjusting the weights and adjustment factors according to specific applications to ensure the reliability and practicality of SI T in different environments.

[0168] SI T The larger the value of SI, the greater the change amplitude of the fluorescence characteristics, indicating a greater change in the activity behavior trend of longhorn beetles;

[0169] Set the median value interval of SI T to [q1, q2]. q1 and q2 are respectively the minimum and maximum values of the median value interval, and [q1, q2] is included in the interval (0, 1);

[0170] Obtain the predicted value N1 of the number of longhorn beetles in the medium exposure mode n(2) ; Denote the warning interval of the number of longhorn beetles on the surface of ancient and famous trees in the current environment as [N1″, N2″]; N1″ and N2″ are respectively the lower limit value and upper limit value of the warning interval of the number of longhorn beetles;

[0171] If the value of SI T is less than q1, and N1 n(2) is within the interval [N1″, N2″], it represents that the change amount of the average fluorescence signal intensity and / or the average fluorescence area in the first time period and the second time period is small, and N1 n(2) within the warning interval [N1″, N2″] indicates that the number of longhorn beetles on the current ancient and famous trees is medium. Therefore, in the scenario where the change amplitude of the fluorescence characteristics is small, the change in the activity behavior trend of longhorn beetles is small, and the number of longhorn beetles is within the warning interval, select the medium exposure mode for shooting; because the low exposure mode will cause underexposure, and the high exposure mode will cause overexposure, and at this time, the medium exposure mode is required to capture the best ultraviolet fluorescence image;

[0172] If the value of SI T is less than q1, and N1 n(2) is less than N1″, it represents that the change amount of the average fluorescence signal intensity and / or the average fluorescence area in the first time period and the second time period is small, and N1 n(2)"Less than N1", indicating that the number of longhorn beetles on the current ancient and famous trees is small. Therefore, in the scenario where the change range of fluorescence characteristics is small, the change trend of the activity behavior of longhorn beetles is small, and the number of longhorn beetles is below the warning range, choosing the low-exposure mode for shooting can handle these small numbers of longhorn beetles;

[0173] If SI T takes a value less than q1, and N1 n(2) is greater than N2", it represents that the change amount of the average fluorescence signal intensity and / or the average fluorescence area of the first time period and the second time period is small, while N1 n(2) is greater than N2", indicating that the number of longhorn beetles on the current ancient and famous trees is large. Therefore, in the scenario where the change range of fluorescence characteristics is small, the change trend of the activity behavior of longhorn beetles is small, and the number of longhorn beetles is above the warning range. Although the change amount of the average fluorescence signal intensity and / or the average fluorescence area of the first time period and the second time period is small, the large number of longhorn beetles themselves will produce relatively large fluorescence characteristic values. Choose to shoot with the three modes of low exposure, medium exposure, and high exposure. Specifically, the high-exposure mode is needed to capture brighter areas. At the same time, it is also necessary to shoot with the three modes of low exposure, medium exposure, and high exposure to avoid detail loss caused by saturated fluorescence intensity;

[0174] If SI T takes a value in the median value interval [q1, q2], and N1 n(2) is within the interval [N1", N2"], it represents that the change amount of the average fluorescence signal intensity and / or the average fluorescence area of the first time period and the second time period is medium, while N1 n(2) is within the warning interval [N1", N2"], indicating that the number of longhorn beetles on the current ancient and famous trees is medium. Therefore, choose to shoot with the low-exposure and medium-exposure modes;

[0175] If SI T takes a value in the median value interval [q1, q2], and N1 n(2) is less than N1", it represents that the change amount of the average fluorescence signal intensity and / or the average fluorescence area of the first time period and the second time period is medium, while N1 n(2) is less than N1", indicating that the number of longhorn beetles on the current ancient and famous trees is small. Therefore, in the case of medium change range of fluorescence characteristics, the change trend of the activity behavior of longhorn beetles is a medium trend. Choose the low-exposure mode for shooting;

[0176] If SI T takes a value in the median value interval [q1, q2], and N1 n(2) is greater than N2", it represents that the change amount of the average fluorescence signal intensity and / or the average fluorescence area of the first time period and the second time period is medium, while N1 n(2)"Greater than N2″ indicates a large number of longhorn beetles on the current ancient and famous trees. Therefore, when the change range of fluorescence characteristics is medium and the change degree of the activity behavior trend of longhorn beetles is medium, three modes of low exposure, medium exposure, and high exposure are selected for joint shooting;

[0177] If SI T takes a value greater than q2, representing a large change in the activity behavior trend of longhorn beetles, which in turn indicates a large change range of fluorescence characteristics, and N1 n(2) is within the interval [N1″, N2″], it means that the number of longhorn beetles on the current ancient and famous trees is within the warning interval, and the change in the activity behavior trend of longhorn beetles is large. Although the number of longhorn beetles does not exceed the warning interval, the change in its activity behavior trend is large. Therefore, it is necessary to select medium exposure and high exposure modes for shooting to avoid detail loss caused by saturated fluorescence intensity;

[0178] If SI T takes a value greater than q2, and N1 n(2) is less than N1″, it means that the change in the activity behavior trend of longhorn beetles is large, but the number of longhorn beetles is in a low quantity state. Since it is difficult for a small number of longhorn beetles to produce more fluorescence substance activity trajectories on the surface of ancient trees, only the medium exposure mode needs to be selected to deal with the large change in the activity behavior trend and the low number of longhorn beetles. The low exposure mode will be underexposed, and the high exposure mode will be overexposed. At this time, the medium exposure mode is needed to capture the best ultraviolet fluorescence image;

[0179] If SI T takes a value greater than q2, and N1 n(2) is greater than N2″, because a large number of longhorn beetles themselves will produce large fluorescence characteristic values, and the change trend of longhorn beetle activities is large, so three modes of low exposure, medium exposure, and high exposure are selected for joint shooting.

[0180] The determination method of [q1, q2] is as follows:

[0181] Method 1, statistical quantile method:

[0182] According to the distribution of SI T determine [q1, q2] through statistical analysis of historical sample data.

[0183] 1. Statistical SI T sample data:

[0184] Collect the SI T change samples in historical longhorn beetle monitoring data, and calculate two appropriate quantile values in the distribution (such as the 25% and 75% quantiles).

[0185] 2. Threshold definition:

[0186] q1 = 25% quantile: corresponding to SI TLower 25% boundary value in the sample;

[0187] q2 = 75th percentile: corresponding to SI T Higher 75% boundary value in the sample.

[0188] With the following characteristics:

[0189] Low-value SI T <q1 corresponds to the lower interval in the sample;

[0190] Medium-value q1 ≤ SI T ≤ q2 corresponds to the median value in the sample set;

[0191] High-value SI T >q2 is an anomaly or concentrated high activity.

[0192] 3. Example:

[0193] Suppose the collected SI T data is [0.1, 0.15, 0.2, 0.3, 0.5, 0.7, 0.9];

[0194] Then the 25th percentile is q1 = 0.2, and the 75th percentile is q2 = 0.7.

[0195] Finally: [q1, q2] = [0.2, 0.7].

[0196] Method 2, empirical definition method referring to actual business requirements:

[0197] Based on the longhorn beetle monitoring requirements and empirical observations of SI T fluctuations to directly set [q1, q2].

[0198] 1. Setting principle:

[0199] According to the system's sensitivity requirements for longhorn beetle behavior:

[0200] If more attention is paid to capturing changes in longhorn beetle activities, set the interval [q1, q2] relatively narrow;

[0201] If attention is paid to system stability and loose monitoring, set the interval [q1, q2] wider.

[0202] Ensure that [q1, q2] is within (0, 1) and conforms to the actual fluctuation situation.

[0203] 2. General empirical values:

[0204] If the detection requirements for diffusion and cluster changes are average, set [q1, q2] = [0.3, 0.7].

[0205] If the impact of longhorn beetle activities on the environment is considered high and key monitoring is required, the interval can be narrowed, such as [0.4, 0.6].

[0206] 3. Adjustment strategy:

[0207] Dynamically adjust [q1, q2] according to on-site conditions (such as environmental light, health status of ancient trees, etc.) to better adapt to the actual situation.

[0208] Method 3: Dynamic adaptive method (dynamically calculated based on real-time monitoring data):

[0209] In the absence of clear statistical data basis or real-time application scenarios, the weight sum formula can be used to dynamically determine [q1, q2].

[0210] 1. Data window based on time series:

[0211] Set a time window (such as the last N2 SI T values), and calculate the mean μ3 and standard deviation σ3 of SI T within this window.

[0212] Dynamically calculate [q1, q2]:

[0213] q1 = μ3 - k3·σ3: It represents the position at k3 times the standard deviation below the mean;

[0214] q2 = μ3 + k3·σ3: It represents the position at k3 times the standard deviation above the mean.

[0215] The recommended range of k3 values is [0.5, 1], which is adjusted according to real-time monitoring requirements.

[0216] The adaptive mechanism of Method 3 can dynamically capture the change trend of SI T and adapt to different sampling stages or monitoring environments.

[0217] Example 2:

[0218] Please refer to Figure 2 , a monitoring system for longhorn beetles on ancient and famous trees based on computer vision, which is used to execute the monitoring method for longhorn beetles on ancient and famous trees based on computer vision, including:

[0219] Sequential image acquisition module: used to continuously acquire ultraviolet fluorescence images generated by the activities of longhorn beetles on the surface of current ancient and famous trees using the medium exposure mode within multiple exposure modes at night or in low-light environments, and record the acquired ultraviolet fluorescence images as sequential image data;

[0220] Preprocessing module: used to preprocess the ultraviolet fluorescence image data in the sequential image data;

[0221] Fluorescent region determination module: It is used to perform semantic segmentation on the preprocessed ultraviolet fluorescent image by using a deep learning model, and extract the fluorescent regions related to longhorn beetle excrement and activity traces;

[0222] Extraction module: It is used to obtain the segmented fluorescent regions and perform extraction and analysis of fluorescent features. The fluorescent features include the area of the fluorescent region and the average fluorescent signal intensity;

[0223] Association model construction module: It is used to obtain the historical data of the number of longhorn beetles under different fluorescent features, and establish an association model between the fluorescent features of the ultraviolet fluorescent image and the number of longhorn beetles according to the fluorescent features of the fluorescent region;

[0224] Extract the fluorescent features of the latest ultraviolet fluorescent image in the time-series image data; input the extracted fluorescent features into the established association model, and finally obtain the predicted value of the number of longhorn beetles based on the medium exposure mode;

[0225] Judgment and adjustment module: It is used to perform trend analysis of the fluorescent features of the currently collected time-series image data for a continuous time period to generate a fluorescent feature change index, and perform combined analysis on the fluorescent feature change index and the predicted value of the number of longhorn beetles. The analysis results are used to make corresponding adjustments to the exposure mode of the subsequent ultraviolet fluorescent images.

[0226] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data and performing software simulation to get a formula closest to the actual situation. The preset parameters in the formulas are set by those skilled in the art according to the actual situation.

[0227] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any other combination. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. Those skilled in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or by a combination of computer software and electronic hardware. Whether these functions are executed by hardware or software methods depends on the specific application and design constraints of the technical solution.

[0228] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units. They may be located in one place, or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0229] As described above, it is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application.

Claims

1. A computer vision-based method for monitoring longhorn beetles on ancient and famous trees, characterized in that: The specific steps include: Step S1: at night or in a low-light environment, initially using a medium exposure mode among multiple exposure modes to continuously collect ultraviolet fluorescence images of the surface of the current ancient and famous tree due to the activity of longhorn beetles, and recording the collected ultraviolet fluorescence images as time-series image data; Step S2: preprocessing the ultraviolet fluorescence image data in the time series image data; Step S3: using a deep learning model to perform semantic segmentation on the preprocessed ultraviolet fluorescence image to extract fluorescent areas related to bovine excrement and activity traces; Step S4: obtaining the segmented fluorescence region and extracting and analyzing the fluorescence features, where the fluorescence features include the fluorescence region area and the average fluorescence signal intensity; Step S5: acquiring historical data of the number of longhorn beetles under different fluorescence characteristics, and establishing a correlation model between the fluorescence characteristics of the ultraviolet fluorescence image and the number of longhorn beetles according to the fluorescence characteristics of the fluorescence area; Extract the fluorescence features of the latest ultraviolet fluorescence image in the time series image data; input the extracted fluorescence features into the established correlation model, and finally obtain the prediction value of the number of longhorn beetles based on the medium exposure mode; Step S6: Perform fluorescence characteristic trend analysis of continuous time periods on the currently acquired time series image data to generate a fluorescence characteristic change index, and combine the fluorescence characteristic change index with the predicted value of the number of longhorn beetles for analysis. The analysis results are used to adjust the exposure mode of subsequent ultraviolet fluorescence images accordingly.

2. The computer vision-based method for monitoring longhorn beetles on ancient and famous trees according to claim 1, characterized in that: The acquisition of the time series image data includes: Using the ultraviolet fluorescence excitation imaging system, multiple ultraviolet fluorescence cameras are arranged around the ancient trees and famous trees, and then continuous image sampling is performed on the surface of the ancient trees and famous trees to obtain a complete ultraviolet fluorescence image of the surface of the ancient trees and famous trees; The multiple exposure modes include: low exposure, medium exposure and high exposure; The exposure time ranges of the multiple exposure modes are set within the interval [BG1, BG2], where BG1 and BG2 are the minimum exposure time and the maximum exposure time, respectively, and the exposure times of low exposure, medium exposure and high exposure increase in sequence; Set i∈{1,2,…,m}, i represents the index of the number of acquisitions, m represents the total number of acquisitions of the ultraviolet fluorescence image in the time series image data, j∈{1,2,3}, j represents the type index of the exposure mode, and j=1 represents the low exposure mode; j=2 represents the medium exposure mode; j=3 represents the high exposure mode; in each acquisition, the initial setting is to perform the ultraviolet fluorescence image acquisition in the medium exposure mode; In the time series image data, the ultraviolet fluorescence image acquired in the jth exposure mode at the i-th time is recorded as i(j); The number of acquisitions in {1,2,…,m} is divided into two parts, the first part represents the number of acquisitions in the first time period, and the second part represents the number of acquisitions in the second time period, and the first time period and the second time period constitute a continuous time period.

3. The computer vision-based method for monitoring longhorn beetles on ancient and famous trees according to claim 2, characterized in that: The preprocessing of the ultraviolet fluorescence image data specifically includes: Based on the bilateral filtering algorithm, low-frequency noise in UV fluorescence images is removed, while high-frequency edge features are enhanced; Histogram equalization was performed on the denoised UV fluorescence image to enhance the contrast between the fluorescent area and the background.

4. The computer vision-based method for monitoring longhorn beetles on ancient and famous trees according to claim 3, characterized in that: The deep learning model performs pixel-level image analysis through the semantic segmentation model UNet, in which the fluorescent areas are classified using pre-trained model parameters; The fluorescent areas in the segmented results are marked by classification masks to distinguish the fluorescent areas from the non-fluorescent areas, where the fluorescent areas include the areas of longhorn beetle excrement and the areas of longhorn beetle activity traces; Generate a multi-channel two-dimensional segmentation mask for the fluorescent area for subsequent fluorescence feature extraction.

5. The computer vision-based method for monitoring longhorn beetles on ancient and famous trees according to claim 4, characterized in that: Define the fluorescence area of ​​the UV fluorescence image i(j) as A f,i(j) ; Based on the area of ​​the fluorescent region, the average fluorescence signal intensity of the fluorescent region is defined as F int,i(j) , the formula is: Where P(x,y) represents the grayscale value of the pixel at the pixel coordinate (x,y); (x,y)∈A f,i(j) Indicates the area A belonging to the fluorescent area f,i(j) The set of all pixel coordinates within.

6. The computer vision-based method for monitoring longhorn beetles on ancient and famous trees according to claim 5, characterized in that: The historical data of the number of longhorn beetles under different fluorescence characteristics were obtained, and the correlation model between the fluorescence characteristics of the ultraviolet fluorescence image and the number of longhorn beetles was established according to the fluorescence characteristics of the fluorescence area, including: At night or in low light, collect a dataset of ultraviolet fluorescence images of ancient and famous trees covering the season when longhorn beetles are active; each ultraviolet fluorescence image records the acquisition time and exposure mode; and also includes records of the number of longhorn beetles corresponding to the manually annotated fluorescence feature data; Match and annotate the fluorescence feature data of the jth exposure mode with the number of longicorn beetles recorded manually in history to generate an annotated data set; Based on the annotated data set, a nonlinear regression model was selected to establish the association model between fluorescence characteristics and the number of longicorn beetles; The association model under the jth exposure mode is defined as: N1 j =h j (F int,i(j) ,A f，i(j) ) Among them, h j is a nonlinear function under the jth exposure mode, which is used to describe the correlation characteristics between the fluorescence characteristics under the jth exposure mode and the number of longicorn beetles; N1 j is the predicted value of the number of longicorns corresponding to the fluorescence characteristics under the jth exposure mode; Extract the fluorescence features of the latest UV fluorescence image in the time series image data, and record the average fluorescence signal intensity and fluorescence area of ​​the fluorescence features extracted under the jth exposure mode in the nth acquisition as F int,n(j) and A f,n(j) ; n represents the acquisition times index of the latest ultraviolet fluorescence image; The F corresponding to the extracted fluorescence characteristic data int,n(j) and A f,n(j) Input into the trained association model, calculate and output the predicted value N1 of the number of longhorn beetles in the jth exposure mode in the nth acquisition n(j) ; The predicted value of the number of longhorn beetles under the medium exposure mode is recorded as N1 n(2) .

7. The computer vision-based method for monitoring longhorn beetles on ancient and famous trees according to claim 6, characterized in that: The first time period and the second time period are respectively denoted as T1 and T2, and T1 and T2 are uniformly represented as T∈{T1, T2}; the fluorescence features in T1 and T2 are differentially calculated to obtain the following average fluorescence signal intensity and average fluorescence area in the first time period and the second time period, respectively: in, and Respectively represent the average fluorescence signal intensity in the first time period and the second time period; and Represent the average fluorescence area in the first time period and the second time period respectively; n T1 Indicates the number of collections corresponding to the first time period; nn T1 Indicates the number of acquisitions corresponding to the second time period; i = n T1 +1 represents the number of first acquisitions in the second time period; The difference between the average fluorescence signal intensity of the second time period and the first time period is calculated to obtain a first difference value ΔF int,T ; The difference between the average fluorescence area of ​​the second time period and the first time period is calculated to obtain the second difference value ΔA f ,T; ΔF int,T and ΔA f ,T calculation formula is as follows: Binding ΔF int,T and ΔA f ,T, define the fluorescence characteristic change index as SI T , the calculation formula is as follows: Among them: α, β are weight coefficients, α+β=1, and the value range of α, β is in the interval (0,1); SI T The larger the value, the greater the change in the fluorescence characteristics, which means the greater the change in the activity trend of the longhorn beetle; Setting SI T The median interval is [q1, q2], q1, q2 are the minimum and maximum values ​​of the median interval respectively, and [q1, q2] is included in the interval (0, 1); the median interval [q1, q2] is used to characterize the boundary range between behaviors with large trend changes and behaviors with small changes; Get the predicted value N1 of the number of longhorn beetles in the medium exposure mode n(2) ; The warning range of the number of longhorn beetles on the surface of ancient trees and famous trees in the current environment is recorded as [N1″, N2″]; N1″ and N2″ are the lower limit and upper limit of the warning range of the number of longhorn beetles respectively; If SI T The value is less than q1, and N1 n(2) When the image is within the range [N1″, N2″], select the medium exposure mode to shoot; If SI T The value is less than q1, and N1 n(2) When the image is smaller than N1″, select low exposure mode to shoot; If SI T The value is less than q1, and N1 n(2) When the image is larger than N2″, choose three modes including low exposure, medium exposure and high exposure to shoot together; If SI T The value is in the median interval [q1, q2], and N1 n(2) When the image is within the range [N1″, N2″], select low exposure and medium exposure modes for shooting; If SI T The value is in the median interval [q1, q2], and N1 n(2) When the image is smaller than N1″, select low exposure mode to shoot; If SI T The value is in the median interval [q1, q2], and N1 n(2) When the image is larger than N2″, choose three modes including low exposure, medium exposure and high exposure to shoot together; If SI T The value is greater than q2, and N1 n(2) When the image is within the range [N1″, N2″], select medium exposure and high exposure modes for shooting; If SI T The value is greater than q2, and N1 n(2) When the image is smaller than N1″, select the medium exposure mode to shoot; If SI T The value is greater than q2, and N1 n(2) When it is larger than N2″, select low exposure, medium exposure and high exposure modes to shoot together.

8. A computer vision-based monitoring system for longhorn beetles on ancient and famous trees, characterized by: The system is used to execute the computer vision-based method for monitoring longhorn beetles on ancient and famous trees according to any one of claims 1 to 7, comprising: Time-series image acquisition module: used to initially use the medium exposure mode in multiple exposure modes to continuously acquire ultraviolet fluorescence images generated by the activities of longicorn beetles on the surface of the current ancient and famous trees at night or in low-light environments, and record the acquired ultraviolet fluorescence images as time-series image data; Preprocessing module: used for preprocessing the ultraviolet fluorescence image data in the time series image data; Fluorescence region determination module: used to perform semantic segmentation on the pre-processed ultraviolet fluorescence image using a deep learning model, and extract the fluorescent regions related to the excretions and activity traces of the cows; Extraction module: used to obtain the segmented fluorescence area and extract and analyze the fluorescence features, including the fluorescence area and the average fluorescence signal intensity; Correlation model building module: used to obtain historical data on the number of longhorn beetles under different fluorescence characteristics, and to establish a correlation model between the fluorescence characteristics of the ultraviolet fluorescence image and the number of longhorn beetles based on the fluorescence characteristics of the fluorescence area; Extract the fluorescence features of the latest ultraviolet fluorescence image in the time series image data; input the extracted fluorescence features into the established correlation model, and finally obtain the prediction value of the number of longhorn beetles based on the medium exposure mode; Judgment and adjustment module: used to perform fluorescence characteristic trend analysis on the currently collected time series image data for continuous time periods to generate a fluorescence characteristic change index, and to conduct a combined analysis of the fluorescence characteristic change index and the predicted value of the number of longicorn beetles. The analysis results are used to make corresponding adjustments to the exposure mode of subsequent ultraviolet fluorescence images.

Images