A biogas slurry overflow monitoring method and system based on intelligent vision

The spill overflow model trained through the YOLO3 algorithm solves the problem of poor spill overflow monitoring effect in the existing technology, realizes efficient identification and timely alarm, and ensures that the environmental hazards of spill overflow are effectively prevented.

CN114202725BActive Publication Date: 2024-12-03MUYUAN FOOD GROUP CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202111484506.0
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2021-12-07
Publication Date
2024-12-03
Estimated Expiration
2041-12-07

AI Technical Summary

Technical Problem

The prior art has poor effect in monitoring of spill overflow, making it difficult to effectively identify spills with irregular shapes and large colors, resulting in untimely pollution monitoring and difficult to prevent environmental hazards.

Method used

The YOLO3 algorithm is used to create a worm fluid spillover model. By collecting reference images of different environments and shapes, and training neural network models, it can quickly identify the on-site images of the worm fluid, and make voice or SMS alarms when spillover is detected.

Benefits of technology

It has achieved efficient identification of spill overflow, with a detection success rate of 98%, and a low false detection rate, ensuring the first-time pollution monitoring and treatment and reducing environmental hazards.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN114202725B_ABST
    Figure CN114202725B_ABST
Patent Text Reader

Abstract

The present invention discloses a biogas slurry overflow monitoring method based on intelligent vision, and the biogas slurry overflow monitoring method comprises the following steps: creating a biogas slurry overflow model for biogas slurry monitoring based on the YOLO3 algorithm; collecting images of a biogas slurry site as images to be monitored; processing the images to be monitored according to the biogas slurry overflow model to determine whether the images to be monitored contain biogas slurry; if the images to be monitored contain biogas slurry, performing a voice alarm and / or a text message alarm. The biogas slurry overflow monitoring method based on intelligent vision provided by the present invention applies the YOLO3 algorithm to biogas slurry overflow monitoring, and can quickly identify whether biogas slurry is contained in the images to be monitored at the biogas slurry site to determine whether there is a biogas slurry overflow phenomenon. Once the biogas slurry overflow is determined, the biogas slurry overflow situation is immediately notified to the relevant responsible person through voice or text message alarm, so as to minimize the environmental harm caused by the biogas slurry overflow risk; and the biogas slurry overflow risk is monitored by the YOLO3 algorithm, with high recognition efficiency and low false detection rate.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of image recognition, and in particular to a biogas slurry overflow monitoring method and system based on intelligent vision. Background Art

[0002] Biogas slurry, also known as manure water, is a mixture of animal feces, urine and water that is stored for a certain period of time and then stirred and fermented. Biogas slurry not only contains abundant macronutrients such as nitrogen, phosphorus, potassium and trace nutrients such as zinc, but also contains 17 kinds of amino acids and active enzymes. These nutrients are basically in the form of fast-acting nutrients. Therefore, biogas slurry has strong fast-acting nutritional capacity and high nutrient availability. It is a multi-dimensional fast-acting compound fertilizer that can be quickly absorbed and utilized by animals and crops.

[0003] However, biogas slurry contains a lot of uric acid, various microorganisms, pollutants, etc. In the process of treating biogas slurry, if it is not handled properly, it is easy to overflow, which will pollute nearby water sources and the environment. If the biogas slurry overflow is monitored by target detection technology, such as performing image recognition on the image after acquiring the image through monitoring, it can be detected whether there is a predetermined target in the image. Due to the characteristics of biogas slurry with irregular shapes, large color differences, and changeable overflow environments, traditional target monitoring methods are all applied to the identification of obvious features (such as animals, plants, vehicles, etc.). When applied to biogas slurry, the monitoring effect is poor. Summary of the invention

[0004] The first purpose of the present invention is to provide a biogas slurry overflow monitoring method based on intelligent vision to solve the problems in the prior art of large biogas slurry overflow pollution and poor effect of traditional target monitoring methods.

[0005] The second objective of the present invention is to provide a biogas slurry overflow monitoring system based on intelligent vision.

[0006] In order to achieve the above object, the present invention provides the following technical solution: a biogas slurry overflow monitoring method based on intelligent vision, the biogas slurry overflow monitoring method comprising the following steps:

[0007] A biogas slurry overflow model for biogas slurry monitoring was created based on the YOLO3 algorithm;

[0008] Collect images of the biogas slurry site as images to be monitored;

[0009] Processing the image to be monitored according to the biogas slurry overflow model to determine whether the image to be monitored contains biogas slurry;

[0010] If the image to be monitored contains biogas slurry, a voice alarm and / or a text message alarm is performed.

[0011] According to an embodiment of the present invention, the creation of a biogas slurry overflow model for biogas slurry monitoring based on the YOLO3 algorithm includes:

[0012] Collect reference images of biogas spills in different environments and shapes;

[0013] Extract target features from the collected reference images of biogas slurry overflow and use them as training sets and test sets;

[0014] The training set is used to train a neural network model based on the YOLO3 algorithm to obtain a biogas slurry overflow model for monitoring.

[0015] According to an embodiment of the present invention, the number of the reference images is 3000-5000.

[0016] According to an embodiment of the present invention, the training set and the test set are divided according to a preset ratio.

[0017] According to an embodiment of the present invention, the preset ratio is divided into 70% and 30%.

[0018] According to an embodiment of the present invention, there are a plurality of images to be monitored, and the images to be monitored are processed by polling according to the biogas slurry overflow model.

[0019] The present invention also provides a biogas slurry overflow monitoring system based on intelligent vision, the biogas slurry overflow monitoring system comprising:

[0020] Model creation module, which creates a biogas slurry overflow model for biogas slurry monitoring based on the YOLO3 algorithm;

[0021] An image acquisition module is provided at the biogas slurry site to acquire images of the biogas slurry site as images to be monitored;

[0022] A control module is configured to import the biogas slurry overflow model and receive the image to be monitored fed back by the image acquisition module, wherein the control module processes the image to be monitored by using the biogas slurry overflow model to determine whether the image to be monitored contains biogas slurry;

[0023] And an alarm module, if the image to be monitored contains biogas slurry, the alarm module performs a voice alarm and / or a text message alarm.

[0024] According to an embodiment of the present invention, the image acquisition module acquires a plurality of images to be monitored, and the control module polls and processes each image to be monitored acquired by the image acquisition module.

[0025] According to an embodiment of the present invention, there are multiple image acquisition modules, and the control module polls and processes the images to be monitored acquired by each of the image acquisition modules.

[0026] According to an embodiment of the present invention, the image acquisition module and the control module are in the same network.

[0027] Compared with the prior art, the biogas slurry overflow monitoring method and system based on intelligent vision provided by the present invention has the following advantages:

[0028] The biogas slurry overflow monitoring method based on intelligent vision provided by the present invention applies the YOLO3 algorithm to the biogas slurry overflow monitoring. Even if the biogas slurry has the characteristics of irregular shape, large color difference and changeable overflow environment, it can quickly identify whether the biogas slurry is contained in the image to be monitored at the biogas slurry site to determine whether there is a biogas slurry overflow phenomenon. Once the biogas slurry overflow is determined, the biogas slurry overflow situation is immediately notified to the relevant responsible person through voice or text message alarm, ensuring that the biogas slurry overflow situation can be resolved in the first time, and the environmental harm caused by the biogas slurry overflow risk is minimized; and the biogas slurry overflow risk is monitored by the YOLO3 algorithm, the recognition efficiency is high, the detection success rate under experimental conditions reaches 98%, and the false detection rate is low. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only used for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. In the accompanying drawings:

[0030] Figure 1 A schematic diagram of a flow chart of a biogas slurry overflow monitoring method based on intelligent vision according to a preferred embodiment of the present invention;

[0031] Figure 2 for Figure 1 The schematic diagram of the process of step S1 in the method for monitoring biogas slurry overflow is shown;

[0032] Figure 3 A schematic structural diagram of a biogas slurry overflow monitoring system based on intelligent vision according to a preferred embodiment of the present invention.

[0033] Reference numerals:

[0034] 1. Model creation module, 2. Image acquisition module, 3. Control module, 4. Alarm module. DETAILED DESCRIPTION

[0035] The present invention provides many applicable creative concepts, which can be embodied in a large number of specific contexts. The specific embodiments described in the following embodiments of the present invention are only exemplary of the specific embodiments of the present invention and do not constitute a limitation on the scope of the present invention.

[0036] First, some proper nouns are explained.

[0037] Object detection: also called object extraction, is an image segmentation based on the geometric and statistical features of the object.

[0038] Intelligent vision: It is a rapidly developing branch of artificial intelligence. Simply put, intelligent vision is to use machines to replace human eyes to make measurements and judgments. For example, the image acquisition mechanism converts the captured target into an image signal, which is transmitted to a dedicated image processing system to obtain the morphological information of the captured target, and then converts it into a digital signal based on the pixel distribution, brightness, color and other information; the image system performs various operations on these signals to extract the characteristics of the target, and then controls the scene based on the judgment results.

[0039] Deep learning: Deep learning is the process of learning the inherent laws and representation levels of sample data. The information obtained in the learning process is very helpful for interpreting data such as text, images and sounds. Its ultimate goal is to enable machines to have analytical learning capabilities like humans and to recognize data such as text, images and sounds. Deep learning is a complex machine learning algorithm that has achieved results in speech and image recognition that far exceed previous related technologies.

[0040] YOLO3 algorithm: It is the third version of the YOLO (You Only Look Once) algorithm. As a method of target detection in the field of intelligent vision, it uses neural networks to directly predict object boundaries and category probabilities, and quickly and accurately achieve end-to-end object monitoring.

[0041] The model design idea is: generate a series of candidate areas on the picture according to certain rules, and then annotate the candidate areas according to the positional relationship between these candidate areas and the real frames of objects on the picture. Those candidate areas that are close enough to the real frames will be marked as positive samples, and the position of the real frames will be used as the position target of the positive samples. Those candidate areas that deviate greatly from the real frames will be marked as negative samples, and negative samples do not need to predict the position or category; use convolutional neural networks to extract image features and predict the position and category of candidate areas, so that each predicted box can be regarded as a sample, and the label value is obtained by annotating the position and category of the real frame relative to it. The position and category are predicted by the network model, and the network prediction value is compared with the label value. Then, a loss function can be established to start the end-to-end training process.

[0042] Convolutional Neural Network: Convolutional neural network is constructed by imitating the visual perception mechanism of biological organisms. It can perform supervised learning and unsupervised learning. The convolution kernel parameter sharing within the hidden layer and the sparsity of the inter-layer connection enable the convolutional neural network to learn grid-like topology features such as pixels and audio with a small amount of computation, with stable results and no additional feature engineering requirements for the data.

[0043] Convolutional neural networks are mainly composed of input layers, convolutional layers, pooling layers, fully connected layers, and output layers. The input layer inputs the training image. The convolutional layer is the core of the convolutional neural network, which is generally composed of multiple convolutional kernels. The convolutional kernel can be understood as a feature extractor. A single convolutional kernel is very small and can only extract local features of the image. However, by increasing the number of convolutional kernels, the convolutional layer can extract multiple local features of the image, and after feature extraction, the convolutional layer can be used for recognition anywhere else in the image. Usually, a convolutional neural network contains multiple convolutional layers. The convolutional layer close to the input layer learns smaller local features, and the next convolutional layer learns larger features of the image based on the features of the previous convolutional layer. The main function of the pooling layer is to reduce the number of model parameters by downsampling without affecting the model effect. The fully connected layer extracts the overall features of the image from the convolutional layer, and the output layer outputs the final result of the model.

[0044] The present invention will be further described below in conjunction with the accompanying drawings and specific implementation methods.

[0045] like Figure 1 As shown, Figure 1 A schematic flow chart of a biogas slurry overflow monitoring method based on intelligent vision according to a preferred embodiment of the present invention.

[0046] This embodiment provides a biogas slurry overflow monitoring method based on intelligent vision, and the biogas slurry overflow monitoring method comprises the following steps:

[0047] S1. Create a biogas slurry overflow model for biogas slurry monitoring based on the YOLO3 algorithm;

[0048] S2, collecting images of the biogas slurry site as images to be monitored;

[0049] S3, processing the image to be monitored according to the biogas slurry overflow model to determine whether the image to be monitored contains biogas slurry;

[0050] S4. If the image to be monitored contains biogas slurry, a voice alarm and / or a text message alarm is issued.

[0051] The intelligent vision-based biogas slurry overflow monitoring method provided in the present application applies the YOLO3 algorithm to the biogas slurry overflow monitoring. Even if the biogas slurry has the characteristics of irregular shape, large color difference, and changeable overflow environment, it can quickly identify whether the image to be monitored at the biogas slurry site contains biogas slurry to determine whether there is a biogas slurry overflow. Once the biogas slurry overflow is determined, the biogas slurry overflow situation is immediately notified to the relevant responsible person through voice or SMS alarm to ensure that the biogas slurry overflow situation can be resolved as soon as possible and the environmental harm caused by the biogas slurry overflow risk is minimized; and the biogas slurry overflow risk is monitored by the YOLO3 algorithm, with high recognition efficiency. The detection success rate under experimental conditions reaches 98%, and the false detection rate is low.

[0052] like Figure 2 As shown, Figure 2 for Figure 1 FIG. 1 is a schematic flow chart of step S1 in the method for monitoring biogas slurry overflow.

[0053] According to an embodiment of the present invention, the step S1 of creating a biogas slurry overflow model for biogas slurry monitoring based on the YOLO3 algorithm comprises:

[0054] S11, collecting reference images of biogas spills in different environments and shapes;

[0055] S12, extracting target features based on the collected reference images of biogas slurry overflow and using them as training sets and test sets;

[0056] S13, using the training set to train a neural network model based on the YOLO3 algorithm to obtain a biogas slurry overflow model for monitoring.

[0057] In this embodiment, the YOLO3 algorithm uses a convolutional neural network with 53 fully connected convolutional layers and uses independent logistic regression for binary classification, so as to extract and mark the location of the biogas slurry in the image.

[0058] According to an embodiment of the present invention, in step S11, the number of reference images is 3000-5000, preferably 4000-5000, for best results. If enough reference images are collected, a sufficient data set will be generated, and the created model will have a wider range of applicability, and will be more accurate in identifying the images to be monitored at the biogas slurry site.

[0059] Specifically, when creating a model, first install Python, NumPy, Pillow, Matplotlib, OpenCV, and TensorFlow environments; among them, Python provides efficient high-level data structures, simple and effective object-oriented programming, Python syntax and dynamic types, as well as the nature of interpreted languages, making it a programming language for writing scripts and rapidly developing applications on most platforms. NumPy (Numerical Python) is an open source numerical computing extension for Python. This tool can be used to store and process large matrices, which is much more efficient than Python's own nested list structure (which can also be used to represent matrices), and supports a large number of dimensional arrays and matrix operations. Pillow is an alternative version of PIL, and the PIL package provides basic image processing functions. Matplotlib is a Python 2D drawing library that generates publication-quality graphics in various hardcopy formats and cross-platform interactive environments. OpenCV is a cross-platform computer vision and machine learning software library released under the Apache 2.0 license (open source), which can run on Linux, Windows, Android and Mac OS operating systems; it is lightweight and efficient - consisting of a series of C functions and a small number of C++ classes, while providing interfaces for languages ​​such as Python, Ruby, and MATLAB, and implementing many common algorithms in image processing and computer vision. TensorFlow is a symbolic mathematical system based on dataflow programming, which is widely used in the programming implementation of various machine learning algorithms. Its predecessor is Google's neural network algorithm library DistBelief.

[0060] Then, download the YOLO3 source code from GitHub, which is a hosting platform for open source and private software projects. Then collect photos of biogas spills in different environments and shapes as reference images, annotate the reference images using labelImg, and save them as XML files; LabelImg is a visual image calibration tool. The datasets required by target detection networks such as Faster R-CNN, YOLO, and SSD all require this tool to calibrate the targets in the image.

[0061] Then, the original image and standard file of the dataset are saved in the VOC directory according to the VOC format; the index txt is generated for the VOC dataset and saved in the VOC directory; the reference image address and standard box position of the sludge overflow are saved as indexes; the .weights file in the YOLO3 algorithm is converted into a .h5 file of Kersa; the anchor prior value of the training data is calculated and the YOLO3 anchor value is updated; the YOLO3 algorithm is trained and saved as a .h5 model.

[0062] According to an embodiment of the present invention, the training set and the test set are divided according to a preset ratio.

[0063] According to an embodiment of the present invention, the preset ratio is divided according to a ratio of 70% and 30%. Of course, in other embodiments, the preset ratio can also be divided according to other ratios.

[0064] In step S2, the biogas slurry site is a factory area containing biogas slurry, and an image acquisition module, such as a surveillance camera, is installed around the biogas slurry site that needs to be monitored. In this embodiment, the image acquisition module is a 4-megapixel camera. Of course, in other embodiments, the image acquisition module can also be other structures. If it is a camera, it can also be other pixels.

[0065] The image acquisition module may be one or more, and is determined according to the direction in which the biogas slurry may overflow.

[0066] In step S3, the biogas slurry overflow model is imported into the control module and the image to be monitored fed back by the image acquisition module is received. The control module processes the image to be monitored through the biogas slurry overflow model to determine whether the image to be monitored contains biogas slurry.

[0067] Specifically, the control module is a Raspberry Pi microcontroller. Raspberry Pi (Chinese name: "Raspberry Pi", abbreviated as RPi, (or RasPi / RPI) is a microcomputer designed for learning computer programming education. It is only the size of a credit card and its system is based on Linux.

[0068] The biogas slurry overflow model is copied to the Raspberry Pi single-chip computer, and a program is written to load the h5 file; a shell file is added, and the Raspberry Pi is set to automatically run the biogas slurry overflow monitoring program when it is turned on.

[0069] In this embodiment, the Raspberry Pi MCU receives the image to be monitored fed back by the surveillance camera, performs image recognition, calls the image to be monitored fed back by other surveillance cameras after 30 seconds, performs image recognition again, and polls and processes all the images to be monitored.

[0070] According to one embodiment of the present invention, there are several images to be monitored, and the images to be monitored are processed by polling according to the biogas slurry overflow model. Polling is a way for the CPU to decide how to provide services to peripheral devices, also known as "programmed I / O". The concept of polling is: the CPU sends out inquiries at regular intervals, asking each peripheral device in sequence whether its service is needed, and providing services if so. After the service is completed, the next peripheral is asked, and then the cycle repeats continuously. The polling processing of the images to be monitored ensures that all locations near the biogas slurry site are monitored, with a wider coverage area, thereby improving the accuracy of biogas slurry monitoring.

[0071] According to an embodiment of the present invention, the image acquisition module and the control module are in the same network. The control module is connected to the image acquisition module according to information for configuring the image acquisition module, such as an IP address and a password.

[0072] In step S4, if the image to be monitored contains biogas slurry, a voice alarm and / or a text message alarm is performed through the alarm module; if the image to be monitored does not contain biogas slurry, step S2 is repeated.

[0073] Specifically, according to the intelligent hierarchical processing of the risk of spillover, a program can be written in the control module to call Alibaba voice service and SMS service. When the biogas slurry spills, a designated person, such as the on-site management post, is called to report the incident through Alibaba voice phone. At the same time, an image of the spilled biogas slurry is intercepted and sent to the designated person via SMS to report the incident, so that relevant personnel can go to the site to shut down the equipment and deal with the spilled biogas slurry, and restart the equipment after the treatment. Of course, in other embodiments, local alarms can also be performed, and the alarm information can be uploaded to the relevant platform.

[0074] like Figure 3 As shown, Figure 3 A schematic structural diagram of a biogas slurry overflow monitoring system based on intelligent vision according to a preferred embodiment of the present invention.

[0075] The present invention also provides a biogas slurry overflow monitoring system based on intelligent vision, the biogas slurry overflow monitoring system comprising:

[0076] Model creation module 1, creating a biogas slurry overflow model for biogas slurry monitoring based on the YOLO3 algorithm;

[0077] An image acquisition module 2 is provided at the biogas slurry site and acquires images of the biogas slurry site as images to be monitored;

[0078] The control module 3 imports the biogas slurry overflow model and receives the image to be monitored fed back by the image acquisition module 2, and the control module 3 processes the image to be monitored through the biogas slurry overflow model to determine whether the image to be monitored contains biogas slurry;

[0079] And an alarm module 4, if the image to be monitored contains biogas slurry, the alarm module performs a voice alarm and / or a text message alarm.

[0080] The model creation module 1 , the image acquisition module 2 , and the alarm module 4 are all connected to the control module 3 .

[0081] The relevant functions and contents of each module have been introduced above and will not be repeated here.

[0082] The image acquisition module 2 is specifically a monitoring camera, and the control module 3 is specifically a Raspberry Pi single-chip microcomputer. The Raspberry Pi single-chip microcomputer is used as an edge detection device, and the monitoring camera is used to obtain the monitored image of the overflowed biogas slurry, which is highly efficient and low in cost.

[0083] According to an embodiment of the present invention, the image acquisition module 2 acquires a plurality of images to be monitored, and the control module 3 polls and processes each image to be monitored acquired by the image acquisition module 2 .

[0084] According to an embodiment of the present invention, there are multiple image acquisition modules 2 , and the control module 3 polls and processes the images to be monitored acquired by each of the image acquisition modules 2 .

[0085] In the above description of this specification, unless otherwise clearly specified and limited, the terms "fixed", "installed", "connected", "connected" and the like should be understood in a broad sense. For example, with regard to the term "connection", it can be a fixed connection, a detachable connection, or an integral one; it can be a mechanical connection or an electrical connection; it can be a direct connection or an indirect connection through an intermediate medium, or it can be the internal connection of two elements or the interaction relationship between two elements. Therefore, unless otherwise clearly defined in this specification, those skilled in the art can understand the specific meanings of the above terms in the present invention according to the specific circumstances.

[0086] According to the above description of this specification, those skilled in the art may also understand that the terms used below, such as "up", "down", "front", "back", "left", "right", "length", "width", "thickness", "vertical", "horizontal", "top", "bottom", "inside", "outside", "axial", "radial", "circumferential", "center", "longitudinal", "lateral", "clockwise" or "counterclockwise", etc., which indicate the orientation or position relationship, are based on the orientation or position relationship shown in the drawings of this specification, and are only for the purpose of facilitating the explanation of the scheme of the present invention and simplifying the description, rather than explicitly or implicitly indicating that the device or element involved must have the specific orientation, be constructed and operate in a specific orientation. Therefore, the above-mentioned orientation or position relationship terms cannot be understood or interpreted as limitations on the scheme of the present invention.

[0087] In addition, the terms "first" or "second" used in this specification to refer to numbers or ordinals are only used for descriptive purposes and cannot be understood as explicitly or implicitly indicating relative importance or implicitly indicating the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of this specification, the meaning of "plurality" is at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0088] As used in this specification and claims, the term "if" may be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting," depending on the context. Similarly, the phrases "if it is determined" or "if [described condition or event] is detected" may be interpreted as "upon determination" or "in response to determining" or "upon detection of [described condition or event]" or "in response to detecting [described condition or event]," depending on the context.

[0089] It should be noted that the above embodiments illustrate rather than limit the present invention, and that those skilled in the art may devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference symbol between brackets shall not be construed as limiting the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims.

Claims

1. A biogas slurry overflow monitoring method based on intelligent vision, characterized in that: The biogas slurry overflow monitoring method comprises the following steps: A biogas slurry overflow model for biogas slurry monitoring was created based on the YOLO3 algorithm; Collect images of the biogas slurry site as images to be monitored; Processing the image to be monitored according to the biogas slurry overflow model to determine whether the image to be monitored contains biogas slurry; If the image to be monitored contains biogas slurry, a voice alarm and / or a text message alarm is issued; The biogas slurry overflow model for biogas slurry monitoring based on the YOLO3 algorithm includes: Collect reference images of biogas spills in different environments and shapes; Extract target features from the collected reference images of biogas slurry overflow and use them as training sets and test sets; The training set is used to train a neural network model based on the YOLO3 algorithm to obtain a biogas slurry overflow model for monitoring.

2. The method for monitoring biogas slurry overflow based on intelligent vision according to claim 1, characterized in that: The number of the reference images is 3000-5000.

3. The method for monitoring biogas slurry overflow based on intelligent vision according to claim 1, characterized in that: The training set and the test set are divided according to a preset ratio.

4. The method for monitoring biogas slurry overflow based on intelligent vision according to claim 3 is characterized in that: The preset ratio is divided into 70% and 30%.

5. The method for monitoring biogas slurry overflow based on intelligent vision according to claim 1, characterized in that: There are a plurality of images to be monitored, and the images to be monitored are processed by polling according to the biogas slurry overflow model.

6. A biogas slurry overflow monitoring system based on intelligent vision, characterized in that: The biogas slurry overflow monitoring system comprises: Model creation module (1), creating a biogas slurry overflow model for biogas slurry monitoring based on the YOLO3 algorithm; An image acquisition module (2) is provided at the biogas slurry site and acquires images of the biogas slurry site as images to be monitored; A control module (3) is configured to import the biogas slurry overflow model and receive the image to be monitored fed back by the image acquisition module (2), wherein the control module (3) processes the image to be monitored using the biogas slurry overflow model to determine whether the image to be monitored contains biogas slurry; and an alarm module (4), wherein if the image to be monitored contains biogas slurry, the alarm module (4) performs a voice alarm and / or a text message alarm; The biogas slurry overflow model for biogas slurry monitoring based on the YOLO3 algorithm includes: Collect reference images of biogas spills in different environments and shapes; Extract target features from the collected reference images of biogas slurry overflow and use them as training sets and test sets; The training set is used to train a neural network model based on the YOLO3 algorithm to obtain a biogas slurry overflow model for monitoring.

7. The biogas slurry overflow monitoring system based on intelligent vision according to claim 6 is characterized in that: The image acquisition module acquires a plurality of images to be monitored, and the control module polls and processes each image to be monitored acquired by the image acquisition module.

8. The biogas slurry overflow monitoring system based on intelligent vision according to claim 6 is characterized in that: There are several image acquisition modules, and the control module polls and processes the images to be monitored acquired by each of the image acquisition modules.

9. The biogas slurry overflow monitoring system based on intelligent vision according to claim 6, characterized in that: The image acquisition module and the control module are in the same network.