Spectrum system based on Web control

Abstract

The invention discloses a spectrum system based on Web control, and relates to the technical field of remote sensing. The problem that real-time spectrum display is difficult to realize on a Web interface is solved; secondly, the spectrum system is difficult to control on a Web interface; then, the category labels of the ground objects are difficult to identify and identify in real time according to the spectral data and the characteristics of the spectral video images; and finally, the Web control spectrum system is difficult to package into application. According to the invention, on the basis of Web native development, a camera picture, spectral data, a calculation result and the like are transmitted to an interface in real time, so that a user can quickly, simply and conveniently check, control and adjust. According to the web control scheme, a user can easily use the web control scheme on various common equipment by packaging the web control scheme into an application, so that the use scene is greatly expanded, the operation efficiency is improved, and the learning cost is reduced.

Description

Technical Field

[0001] The present invention belongs to the technical field of remote sensing, and specifically relates to a spectral system based on Web control. Background Art

[0002] With the development of remote sensing technology, the acquisition and analysis of spectral information of ground objects have become increasingly important. A ground object spectrometer, also known as a hyperspectral ground object spectrometer, is an instrument for measuring the spectral radiation or reflection characteristics of ground objects. In the visible and near-infrared regions, it mainly measures the spectral reflection characteristics of objects. By irradiating the ground object with a light source, and then receiving and analyzing the reflected spectral information. This information is dispersed into light of different wavelengths by a spectral splitting system, and then the intensity is measured by a detection system, and finally analyzed and calculated by a data processing system to obtain the spectral characteristics of the object. Traditional ground object spectrometers usually need to be connected to a computer by a wired method for data transmission and processing. Such equipment that occupies a large area and is bulky is obviously not convenient for outdoor operation and also has a high learning threshold.

[0003] The following problems exist in the prior art: First, it is difficult to realize real-time display of spectra on a Web interface; second, it is difficult to control the spectral system on a Web interface; third, it is difficult to identify and authenticate the category labels of ground objects in real time according to the characteristics of spectral data and spectral video images; finally, it is difficult to package the Web-controlled spectral system into an application. Summary of the Invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art; for this purpose, the present invention proposes a spectral system based on Web control to solve the following technical problems:

[0005] First, it is difficult to realize real-time display of spectra on a Web interface; second, it is difficult to control the spectral system on a Web interface; third, it is difficult to identify and authenticate the category labels of ground objects in real time according to the characteristics of spectral data and spectral video images; finally, it is difficult to package the Web-controlled spectral system into an application.

[0006] To solve the above problems, the first aspect of the present invention provides a spectral system based on Web control, including the following modules:

[0007] Data acquisition module: Calibrate the ground object spectrometer, and collect spectral data and spectral video images in real time by setting the measurement parameters of the calibrated ground object spectrometer.

[0008] Data transmission module: Send the collected spectral data and spectral video images to the server side through the TCP port.

[0009] Data Processing Module: Preprocess the spectral data received on the server side, including: denoising, smoothing, and calibration; preprocess the received spectral video images, including: denoising and image enhancement;

[0010] Spectral Analysis Module: Draw spectral curves based on the collected and processed spectral data and perform feature extraction according to the spectral curves; perform feature extraction based on the preprocessed spectral video images; identify and classify the category labels of ground objects in real time according to the feature extraction results; analyze and calculate the statistical characteristics of the spectral data;

[0011] Interaction Control Module: Design a Web interface and perform interactive control through the Web interface;

[0012] Packaging and Application Module: Integrate the ground object spectrometer, the front-end Web control interface, and the back-end server, and package the Web control-based spectral system into an application and deploy it to the server.

[0013] As a further solution of the present invention: Real-time collect spectral data and spectral video images by setting the measurement parameters of the ground object spectrometer, including the following steps:

[0014] The measurement parameters include: spectral range, sensitivity, resolution, and sampling time; the spectral data includes: radiance, irradiance, spectral reflectance, and spectral distribution; record the spectral data and spectral video images collected at each time point.

[0015] As a further solution of the present invention: Send the collected spectral data and spectral video images to the server side through the TCP port, including the following steps: The server side creates a Web server using Nginx and PHP to provide static content; forward the TCP port of the main program to HTTP using PHP, and the front end parses the JSON data transmitted from the server side; after the configuration is completed, forward the spectral data and spectral video images collected by the ground object spectrometer from the TCP port to the HTTP port.

[0016] As a further solution of the present invention: Draw spectral curves based on the collected and processed spectral data and perform feature extraction according to the spectral curves, including the following steps:

[0017] Draw spectral curves based on the collected and processed spectral data, where the wavelength is represented on the horizontal axis and the absorbance or reflectance is represented on the vertical axis, and perform denoising and smoothing processing on the spectral curves;

[0018] By analyzing the formula:

[0019] Obtain the first derivative R'(λ) of the reflectance curve; where λ represents the wavelength and R represents the reflectance; when R'(λ) = 0, obtain the potential extreme points, and the potential extreme points include: potential peaks or potential valleys; where the peak represents the wavelength position with the maximum reflectance or the minimum absorbance, and the valley represents the wavelength position with the minimum reflectance or the maximum absorbance;

[0020] Perform a second derivative of R'(λ), and by analyzing the formula:

[0021] Obtain the second derivative R”(λ); when R”(λ) = 0, obtain the inflection point of the reflectance curve; when R”(λ)>0, obtain the valley of the reflectance curve; when R”(λ)<0, obtain the peak of the reflectance curve;

[0022] Perform feature extraction based on the spectral curve, including: reflectance, absorption characteristics, reflection peaks, the peak of the spectral curve, the valley of the spectral curve, bandwidth, curve peak position, peak intensity, peak wavelength, and inflection point.

[0023] As a further solution of the present invention: perform feature extraction based on the preprocessed spectral video image, including the following steps:

[0024] Based on the collected and processed video image, the extracted features include: spatial features, time series features, and motion features; for the spatial features, through image processing techniques, including: edge detection and texture analysis techniques to extract spatial features, including: texture features, shape features, and edge features; for the motion features, through the inter-frame difference method and background subtraction method techniques, extract the motion features of the moving target.

[0025] As a further solution of the present invention: based on the feature extraction results, identify and authenticate the class labels of the ground objects in real time, including the following steps:

[0026] Based on the features extracted from the real-time spectral data and spectral video image, perform feature dimensionality reduction on the real-time extracted features and then use the classification model for identification and authentication;

[0027] Obtain a dataset containing known ground object labels, where each sample's corresponding ground object label and features are recorded; divide the dataset into a training set and a test set; input the training set into the classification model for training;

[0028] Use the test set data to evaluate the performance metrics of the accuracy or recall rate of the classification model; when the performance metrics do not meet the threshold, readjust the model or features;

[0029] When the performance index meets the threshold, based on the trained classification model, feature extraction is performed on the spectral data and spectral video images that are real-time collected and processed by the Web-controlled spectral system; the features extracted in real time are input into the trained classification model after feature dimensionality reduction for real-time identification and the identification result is output; the identification result of the classification model is displayed to the user through the Web interface, and the identification result is the class label of the ground object.

[0030] As a further solution of the present invention: analyzing and calculating the statistical characteristics of the spectral data, including the following steps:

[0031] By analyzing the formula:

[0032] The average value μ of the reflectivity is obtained, where n is the number of wavelength points, and R(λ i ) represents the reflectivity of the i-th wavelength point;

[0033] By analyzing the formula:

[0034] The standard deviation σ of the reflectivity is obtained;

[0035] By analyzing the formula:

[0036] The coefficient of variation CV of the reflectivity is obtained, where σ represents the standard deviation of the reflectivity and μ represents the average value of the reflectivity;

[0037] By analyzing the formula:

[0038] The kurtosis F of the reflectivity is obtained;

[0039] By analyzing the formula:

[0040] The skewness P of the reflectivity is obtained.

[0041] As a further solution of the present invention: designing a Web interface and performing interactive control through the Web interface, including the following steps:

[0042] Designing the Web interface includes: a spectral area, a video area, a parameter area, a control area, and a setting area;

[0043] The camera images, spectral data, and the results of analysis and calculation are transmitted in real time to a Web interface for real-time display; after calibration in the control area on the Web interface, the measurement results are displayed in the spectral area and the video area on the Web interface; the user adjusts parameters in the setting area on the Web interface; the front end sends new parameters to the back end, and the back end then passes them to the main program for update, and the displayed spectral data, spectral video images, and new measurement results are updated in real time; according to the calibration performed by the user in the control area, the measurement parameters are adjusted, and the measurement results are dynamically displayed in the spectral area and the parameter area.

[0044] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0045] Through the feature extraction results of spectral data and spectral video images, the present invention can use a classification model to identify and authenticate the category labels of ground objects in real time; by analyzing the statistical characteristics of spectral data, the spectral characteristics of ground objects can be understood more deeply.

[0046] By designing a simple and clear Web interface, the present invention enables users to operate and control the system more intuitively, reducing the usage difficulty. Through interactive control via the Web interface, the feedback and results of the system can be obtained in real time, improving the operation efficiency.

[0047] By using a web-controlled spectral system encapsulated as an application, users can easily use it on various common devices, such as mobile phones and tablets with HarmonyOS, Android, or iOS systems, or computers with MacOS or Windows systems, greatly expanding the usage scenarios, improving the operation efficiency, and reducing the learning cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required for use in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0049] Figure 1 It is a schematic diagram of the module structure of the present invention;

[0050] Figure 2 It is a schematic diagram of the Web page of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0051] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0052] Please refer to Figure 1 - Figure 2 As shown, the first aspect embodiment of the present invention provides a Web control-based spectral system, including the following modules:

[0053] Data acquisition module: Calibrate the ground object spectrometer, and collect spectral data and spectral video images in real time by setting the measurement parameters of the calibrated ground object spectrometer.

[0054] Data transmission module: Send the collected spectral data and spectral video images to the server side through the TCP port.

[0055] Data processing module: Preprocess the received spectral data on the server side, including: denoising, smoothing and correction; preprocess the received spectral video images, including: denoising and image enhancement.

[0056] Spectral analysis module: Draw a spectral curve based on the collected and processed spectral data and perform feature extraction according to the spectral curve; perform feature extraction according to the preprocessed spectral video images; identify and classify the category labels of ground objects in real time according to the feature extraction results; analyze and calculate the statistical characteristics of the spectral data.

[0057] Interaction control module: Design a Web interface and perform interactive control through the Web interface.

[0058] Packaging and application module: Integrate the ground object spectrometer, the front-end Web control interface and the back-end server together, and package the Web control-based spectral system into an application and deploy it to the server.

[0059] Specifically, a standard whiteboard or a standard substance with a known reflectivity is used to calibrate the ground object spectrometer. Set the measurement parameters of the calibrated ground object spectrometer to meet the acquisition requirements of real-time spectral data and spectral video images. Align the probe of the ground object spectrometer with the ground object to be measured, ensuring that the measurement conditions are consistent, including: illumination, distance, etc. Real-time collect spectral data and spectral video images through the ground object spectrometer. Configure a TCP port on the server side for receiving spectral data and spectral video images from the data acquisition module. Send the collected spectral data and spectral video images to the server side through the TCP port. Perform denoising processing and baseline correction on the received spectral data; perform denoising, enhance the contrast and edge details of the received spectral video images, and highlight the key features in the images. According to the collected and processed spectral data, plot a spectral curve and extract the features of the spectral data from the spectral curve; extract features from the preprocessed spectral video images. According to the feature extraction results, use a classification model to identify and classify the category labels of the ground objects in real time. Analyze the distribution law and change trend of the spectral data by analyzing and calculating statistical characteristics such as the mean value, standard deviation, kurtosis, and skewness of the spectral data. Design a Web interface and allow users to interactively control through the Web interface to implement functions such as users controlling data acquisition, data transmission, data processing, and spectral analysis through the Web interface. Provide real-time feedback and status display to let users understand the operating status and results of the system. Integrate the ground object spectrometer, the front-end Web control interface, and the back-end server, and package the Web-based spectral system into an application, and deploy the packaged application to the server.

[0060] In one embodiment of the present invention, real-time collection of spectral data and spectral video images is performed by setting the measurement parameters of the ground object spectrometer, including the following steps:

[0061] The measurement parameters include: spectral range, sensitivity, resolution, and sampling time; the spectral data includes: radiance, irradiance, spectral reflectance, and spectral distribution; record the spectral data and spectral video images collected at each time point.

[0062] Specifically, ensure that the ground object spectrometer is in good working condition, check whether all components are intact, especially whether the probe and optical fiber are clean and undamaged. Calibrate the ground object spectrometer by using a standard whiteboard or a standard substance with a known reflectivity to ensure the accuracy of the measurement data. Set measurement parameters such as the spectral range, sensitivity, resolution, and sampling time according to the characteristics of the ground object to be measured and the measurement requirements. Usually, it is selected according to the spectral characteristics of the ground object to be measured. For example, set the spectral range from near ultraviolet to near infrared or a wider range; adjust according to the reflectivity or transmittance of the ground object to be measured to ensure that weak optical signals can be detected to set the sensitivity. Select an appropriate resolution according to the measurement requirements to distinguish different spectral features. Adjust according to the measurement time and data volume requirements to ensure that sufficient spectral data and video images can be collected in real time. Align the probe of the ground object spectrometer with the ground object to be measured to ensure that the measurement environmental conditions are consistent. Start the data acquisition function and start collecting spectral data and spectral video images in real time. During the acquisition process, record the spectral data and spectral video images collected at each time point. The spectral data includes, but is not limited to, key information such as radiance, irradiance, spectral reflectance, and spectral distribution. The spectral video images record the dynamic change process of the ground object to be measured.

[0063] In one embodiment of the present invention, sending the collected spectral data and spectral video images to the server side through the TCP port includes the following steps: The server side creates a Web server by using Nginx and PHP to provide static content; forwards the TCP port of the main program to HTTP by using PHP, and the front end parses the JSON data transmitted from the server side; after the configuration is completed, forward the spectral data and spectral video images collected by the ground object spectrometer from the TCP port to the HTTP port.

[0064] Specifically, according to the server operating system type, download and install Nginx, and configure Nginx to provide static content services, including files such as HTML, CSS, JavaScript, etc. Download and install PHP according to the server operating system type. Configure PHP to work in coordination with Nginx to ensure that PHP scripts can be executed correctly. Modify the Nginx configuration file to forward PHP requests to PHP-FPM for processing. Ensure normal communication between Nginx and PHP-FPM so that PHP scripts can be processed. Create a PHP script to receive data on a TCP port and forward it to an HTTP port. Use the socket extension of PHP to create a socket object and listen on the specified TCP port. When a TCP connection is received, read the data and convert it to JSON format. According to the instruction manual or development documentation of the ground object spectrometer, configure its data acquisition and transmission parameters. Set the spectrometer to send data to the specified TCP port on the server side. After configuration, use the ground object spectrometer to conduct data acquisition and transmission tests to ensure that the data can be correctly sent to the TCP port on the server side and received by the PHP script and forwarded to the HTTP port. Receive the JSON data returned by the server and parse it into available spectral data and spectral video image information. Use front-end technology to draw the spectral video image on the page and display the relevant information of the spectral data. The above-mentioned Ngix can be replaced by other common web servers such as Apache or directly integrated into the main program. The PHP program can also be replaced by directly implementing HTTP interaction in the main program. The above-mentioned HTTP can be directly replaced by HTTPS, etc.

[0065] In one embodiment of the present invention, according to the collected and processed spectral data, a spectral curve is drawn and feature extraction is performed based on the spectral curve, including the following steps:

[0066] According to the collected and processed spectral data, a spectral curve is drawn, where the wavelength is represented on the horizontal axis and the absorbance or reflectance is represented on the vertical axis, and the spectral curve is denoised and smoothed.

[0067] By analyzing the formula:

[0068] The first derivative R'(λ) of the reflectance curve is obtained; where λ represents the wavelength and R represents the reflectance; when R'(λ) = 0, potential extreme points are obtained, and the potential extreme points include: potential peaks or potential valleys; where the peak represents the wavelength position with the maximum reflectance or the minimum absorbance, and the valley represents the wavelength position with the minimum reflectance or the maximum absorbance.

[0069] The second derivative of R'(λ) is taken, and by analyzing the formula:

[0070] Obtain the second derivative R”(λ); when R”(λ) = 0, obtain the inflection point of the reflectivity curve; when R”(λ) > 0, obtain the valley value of the reflectivity curve; when R”(λ) < 0, obtain the peak value of the reflectivity curve;

[0071] Perform feature extraction based on the spectral curve, including: reflectivity, absorption characteristics, reflection peak, peak value of the spectral curve, valley value of the spectral curve, bandwidth, curve peak position, peak intensity, peak wavelength, and inflection point.

[0072] Specifically, based on the collected and processed spectral data, plot the spectral curve, with the wavelength represented on the horizontal axis and the absorbance or reflectivity represented on the vertical axis. Plot the spectral curve according to the wavelength and the corresponding absorbance or reflectivity values of the spectral data. Apply a filter to remove high-frequency noise in the spectral data; reduce random fluctuations in the data by using smoothing algorithms including: moving average method or Savitzky-Golay smoothing, etc., to obtain a smoother spectral curve. Use a formula to calculate the first derivative of the reflectivity curve, and when the first derivative is equal to 0, obtain potential extreme points, where the potential extreme points include potential peaks and potential valleys; use the second derivative formula to calculate the second derivative of the reflectivity curve, where the second derivative can reflect the curvature change of the spectral curve, thereby identifying the inflection point. When the second derivative is equal to 0, identify the inflection point of the reflectivity curve. Extract features such as reflectivity, absorption characteristics, reflection peak, peak value of the spectral curve, and valley value of the spectral curve from the spectral curve.

[0073] In one embodiment of the present invention, feature extraction is performed based on the preprocessed spectral video image, including the following steps:

[0074] Based on the collected and processed video image, extract features including: spatial features, time series features, and motion features; for the spatial features, through image processing techniques, including: edge detection and texture analysis techniques to extract spatial features, including: texture features, shape features, and edge features; for the motion features, through the inter-frame difference method and background subtraction method techniques, extract the motion features of the moving target.

[0075] Specifically, grayscale processing is performed on the preprocessed spectral video image, and texture analysis technology is applied to extract the texture features of the image. Edge detection is performed on the preprocessed image to obtain the contour of the object, and according to the contour information, the shape features of the object are extracted, such as perimeter, area, aspect ratio, etc. A suitable edge detection algorithm is selected, and the edge detection algorithm is applied to the preprocessed image to obtain an edge image, and edge features are extracted from the edge image, including: the position, direction, intensity, etc. of the edge. Time series analysis is performed on the preprocessed spectral video image to extract features related to the time series, including: lag features, i.e., new features created based on the data values at previous time points, used to capture the delay effect in the time series, rolling window features, i.e., features extracted by calculating statistics within a specified window, reflecting the local trend of the data, etc. Frame difference operation is performed on the preprocessed spectral video image, and the difference result is binarized to obtain the contour of the moving target; the motion features of the moving target are extracted from the binarized image, including: motion speed, motion direction, etc.

[0076] In one embodiment of the present invention, according to the feature extraction results, the category label of the ground object is identified and authenticated in real time, including the following steps:

[0077] According to the features extracted from the real-time spectral data and spectral video image, the features extracted in real time are subjected to feature dimensionality reduction and then identified and authenticated using a classification model;

[0078] Obtain a data set containing known ground object labels, where the ground object label and features corresponding to each sample are recorded; divide the data set into a training set and a test set; input the training set into the classification model for training;

[0079] Use the test set data to evaluate the performance metrics of the accuracy or recall rate of the classification model; when the performance metrics do not meet the threshold, readjust the model or features;

[0080] When the performance metrics meet the threshold, based on the trained classification model, the spectral data and spectral video image collected and processed in real time by the Web-controlled spectral system are subjected to feature extraction; the features extracted in real time are subjected to feature dimensionality reduction and then input into the trained classification model for real-time identification and authentication, and the identification and authentication results are output; the identification and authentication results of the classification model are displayed to the user through the Web interface, and the identification and authentication results are the category labels of the ground objects.

[0081] Specifically, dimensionality reduction techniques such as principal component analysis and minimum noise separation transformation are used to perform dimensionality reduction on the spectrogram data and spectrogram video image features extracted in real time, obtaining the dimensionality-reduced feature vectors. A data set containing known ground object labels is obtained, where each sample corresponds to a ground object label and features. The data set is divided into a training set and a test set. Usually, the training set accounts for the majority and the test set accounts for a small part. The training set is input into classification models, including support vector machine (SVM), random forest, convolutional neural network (CNN), etc., for training to obtain a trained classification model. The test set data is used to evaluate performance metrics such as the accuracy or recall rate of the classification model. A threshold is set at 96%, and the threshold is dynamically adjusted according to the actual situation; if the performance metrics do not meet the set threshold, the model parameters or feature selection are readjusted until the performance metrics meet the requirements. The spectrogram system based on Web control collects spectrogram data and spectrogram video images in real time; the collected spectrogram data and spectrogram video images are preprocessed and feature extracted to obtain real-time features, the features extracted in real time are subjected to dimensionality reduction processing, and the dimensionality-reduced features are input into the trained classification model. The classification model identifies and authenticates the input features and outputs the category labels of the ground objects. The identification and authentication results of the classification model are displayed to the user through the Web interface, facilitating the user to view and understand the ground object category information in real time.

[0082] In one embodiment of the present invention, the statistical characteristics of the spectrogram data are analyzed and calculated, including the following steps:

[0083] By analyzing the formula:

[0084] The average value μ of the reflectivity is obtained, where n is the number of wavelength points, and R(λ i ) represents the reflectivity of the i-th wavelength point;

[0085] By analyzing the formula:

[0086] The standard deviation σ of the reflectivity is obtained;

[0087] By analyzing the formula:

[0088] The coefficient of variation CV of the reflectivity is obtained, where σ represents the standard deviation of the reflectivity and μ represents the average value of the reflectivity;

[0089] By analyzing the formula:

[0090] The kurtosis F of the reflectivity is obtained;

[0091] By analyzing the formula:

[0092] The skewness P of the reflectivity is obtained.

[0093] Specifically, the reflectance data at all wavelength points are collected, and the statistical characteristics of the spectral data are obtained by using the formula, including: the average value, standard deviation, coefficient of variation, kurtosis, and skewness of the reflectance. Analyzing and calculating the statistics of the spectral data helps to understand the overall distribution pattern and dispersion degree of the reflectance data.

[0094] In one embodiment of the present invention, a Web interface is designed and interactive control is performed through the Web interface, including the following steps:

[0095] Designing the Web interface includes: a spectral area, a video area, a parameter area, a control area, and a setting area;

[0096] The camera screen, spectral data, and analysis calculation results are transmitted in real time to the Web interface for real-time display; after calibration in the control area on the Web interface, the measurement results are displayed in the spectral area and video area on the Web interface; the user adjusts the parameters in the setting area on the Web interface; the front end sends the new parameters to the back end, and the back end then passes them to the main program for update, and the displayed spectral data, spectral video images, and new measurement results are updated in real time; according to the calibration performed by the user in the control area, the measurement parameters are adjusted, and the measurement results are dynamically displayed in the spectral area and parameter area.

[0097] Specifically, after the user clicks the calibration button in the control area, the calibration process is triggered, and during the calibration process, the user may need to cooperate with some operations, such as adjusting the camera position, adjusting the spectrometer parameters, etc. After calibration is completed, the calibrated measurement results are displayed in the spectral area and video area. The user can view the current parameter values in the parameter area and adjust them by means of a slider, input box, etc. The adjusted parameters are sent to the back-end program for update, and the displayed spectral data, spectral video images, and measurement results are updated in real time. The user can adjust the parameters in the setting area. When the user adjusts the parameters or makes setting adjustments, the front end will send the new parameter values to the back-end program for update. This can be achieved through technologies such as AJAX for asynchronous requests and responses. After the back-end program receives the new parameter values, it will pass them to the main program for update. The main program re-measures and analyzes the spectral data according to the new parameter values. The updated spectral data, analysis results, and camera screen of the main program are transmitted in real time to the front-end Web interface for updated display. This ensures that the user can see the latest measurement results and camera screen in real time.

[0098] The above embodiments are only used to illustrate the technical method of the present invention and not to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical method of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A Web-based spectroscopy system, characterized in that: Includes the following modules: Data acquisition module: calibrate the ground object spectrometer, and collect spectral data and spectral video images in real time by setting the measurement parameters of the calibrated ground object spectrometer; Data transmission module: sends the collected spectral data and spectral video images to the server through the TCP port; Data processing module: pre-processing the received spectral data on the server side, including: denoising, smoothing and correction; pre-processing the received spectral video images, including: denoising and image enhancement; Spectral analysis module: based on the collected and processed spectral data, draw spectral curves and perform feature extraction based on the spectral curves; perform feature extraction based on the pre-processed spectral video images; based on the feature extraction results, identify and identify the category labels of the objects in real time; analyze and calculate the statistical characteristics of the spectral data; Interactive control module: design the Web interface and perform interactive control through the Web interface; Encapsulate application module: Integrate the ground object spectrometer, front-end Web control interface and back-end server, and encapsulate the Web-controlled spectral system into an application and deploy it on the server.

2. A Web-controlled spectroscopy system according to claim 1, characterized in that: The spectral data and spectral video images are collected in real time by setting the measurement parameters of the ground object spectrometer, including the following steps: The measurement parameters include: spectral range, sensitivity, resolution and sampling time; the spectral data include: radiant brightness, radiant illumination, spectral reflectance and spectral distribution; the spectral data and spectral video images collected at each time point are recorded.

3. A Web-controlled spectroscopy system according to claim 1, characterized in that: The collected spectral data and spectral video images are sent to the server side through the TCP port, including the following steps: the server side is to create a Web server by using Nginx and PHP to provide static content; the TCP port of the main program is forwarded to HTTP by using PHP, and the front end parses the JSON data sent from the server side; after the configuration is completed, the spectral data and spectral video images collected by the ground object spectrometer are forwarded from the TCP port to the HTTP port.

4. A Web-controlled spectroscopy system according to claim 1, characterized in that: According to the collected and processed spectral data, a spectral curve is drawn and feature extraction is performed according to the spectral curve, including the following steps: According to the collected and processed spectral data, a spectral curve is drawn, wherein the horizontal axis represents the wavelength and the vertical axis represents the absorbance or reflectance, and the spectral curve is denoised and smoothed; By analyzing the formula: Obtaining the first-order derivative R'(λ) of the reflectivity curve; wherein λ represents the wavelength and R represents the reflectivity; when R'(λ)=0, obtaining a potential extreme point, the potential extreme point includes: a potential peak value or a potential valley value; wherein the peak value represents the wavelength position at which the reflectivity is maximum or the absorbance is minimum, and the valley value represents the wavelength position at which the reflectivity is minimum or the absorbance is maximum; Take the second derivative of R'(λ), and analyze the formula: The second-order derivative R" (λ) is obtained; when R" (λ) = 0, the inflection point of the reflectivity curve is obtained; when R" (λ) > 0, the valley value of the reflectivity curve is obtained; when R" (λ) < 0, the peak value of the reflectivity curve is obtained; Feature extraction is performed based on the spectral curve, including: reflectivity, absorption characteristics, reflection peak, peak value of the spectral curve, valley value of the spectral curve, bandwidth, curve peak position, peak intensity, peak wavelength and inflection point.

5. A Web-controlled spectroscopy system according to claim 1, characterized in that: Feature extraction is performed based on the preprocessed spectral video image, including the following steps: According to the collected and processed video images, the extracted features include: spatial features, time series features and motion features; the spatial features are extracted through image processing technology, including: edge detection and texture analysis technology, including: texture features, shape features and edge features; the motion features are extracted through inter-frame difference method and background subtraction method to extract the motion features of the moving target.

6. The Web-controlled spectroscopy system according to claim 1, characterized in that: According to the feature extraction results, the category labels of the objects are identified in real time, including the following steps: According to the features extracted from real-time spectral data and spectral video images, the features extracted in real time are subjected to feature dimension reduction and then identified and evaluated using a classification model; Obtain a data set containing known object labels, in which the object label and features corresponding to each sample are recorded; split the data set into a training set and a test set; input the training set into a classification model for training; Use the test set data to evaluate the performance indicators of the classification model's accuracy or recall; when the performance indicators do not meet the threshold, re-adjust the model or features; When the performance index meets the threshold, feature extraction is performed based on the trained classification model and the spectral data and spectral video images collected and processed in real time by the Web-controlled spectral system; the real-time extracted features are input into the trained classification model after feature dimensionality reduction, and real-time recognition and identification are performed and the recognition and identification results are output; the recognition and identification results of the classification model are displayed to the user through the Web interface, and the recognition and identification results are the category labels of the objects.

7. A Web-controlled spectroscopy system according to claim 1, characterized in that: The statistical characteristics of the spectral data are analyzed and calculated, including the following steps: By analyzing the formula: Get the average reflectivity μ, where n is the number of wavelength points, R(λ i ) represents the reflectivity at the i-th wavelength point; By analyzing the formula: Get the standard deviation σ of the reflectivity; By analyzing the formula: The coefficient of variation CV of the reflectivity is obtained, where σ represents the standard deviation of the reflectivity and μ represents the average value of the reflectivity; By analyzing the formula: Get the kurtosis F of the reflectivity; By analyzing the formula: The skewness P of the reflectivity is obtained.

8. The Web-controlled spectroscopy system according to claim 1, characterized in that: Designing a web interface and performing interactive control through the web interface includes the following steps: Design the web interface to include: spectrum area, video area, parameter area, control area and setting area; The camera images, spectral data and analysis and calculation results are transmitted to the Web interface in real time for real-time display; after calibration in the control area on the Web interface, the spectral area and video area on the Web interface display the measurement results; the user adjusts the parameters in the setting area on the Web interface; the front end sends the new parameters to the back end, and the back end passes them to the main program for update, and the displayed spectral data, spectral video images and New measurement results; adjust measurement parameters according to the calibration performed by the user in the control area, The measurement results are dynamically displayed in the spectrum area and parameter area.

Images