Open slope landslide mass identification and early warning system and method based on infrared local temperature increasing effect

Abstract

The invention discloses an open-air slope landslide mass identification and early warning system and method based on an infrared local temperature increasing effect, and the method comprises the steps: carrying out the non-contact monitoring through an infrared thermal imager, combining a deep learning U-Net network, a VGG network and an Adam algorithm, and carrying out the intelligent identification of an open-air slope landslide mass based on the local temperature increasing characteristics in an infrared thermogram. Specifically, the method comprises the following steps: firstly, acquiring temperature increasing data by using an infrared thermogram, and carrying out data cleaning and slope area delineation through an image processing technology; then, a deep learning model is used for infrared local warming intelligent identification, and a landslide mass activity area is extracted from the deep learning model; and finally, comprehensive analysis and risk assessment are carried out by combining geological data, meteorological data and the like, and a landslide early warning signal is sent out. According to the method, high-precision landslide mass recognition can be achieved, the mPA value is 94.37%, the mIoU value is 89.21%, the accuracy and real-time performance of landslide monitoring are effectively improved, and the method has wide application prospects and is particularly suitable for landslide disaster prevention and control of strip mine areas and high-risk side slopes.

Description

Technical Field

[0001] The present invention relates to the technical field of geological disaster prevention and control, and in particular to an open-air slope landslide body identification and early warning system and method based on infrared local warming effect. Background Art

[0002] Slope landslide is one of the common geological disasters, which seriously threatens the safety of human life and property and has a huge impact on social and economic activities. Early identification and early warning of landslides are the basis of disaster prevention and mitigation work and the key to effectively reducing the losses caused by landslide disasters. However, traditional landslide identification methods often rely on manual surveys, which are inefficient and easily limited by the experience and knowledge level of staff. The situation at the project site is complex and changeable. In addition to the warming phenomenon of the landslide body itself, pedestrians, buildings, water bodies and forests in the surrounding environment may also show local warming phenomena. The warming interference of these non-landslide bodies often leads to misjudgment and distortion, which greatly affects the accuracy of landslide body identification.

[0003] With the development of science and technology, infrared thermal imaging technology, as a non-contact, real-time monitoring method, has gradually been applied to landslide identification research. By monitoring the slope area with an infrared thermal imager, the signal of local warming can be captured before the landslide occurs, thus providing an important basis for early warning of landslides. Although traditional image processing methods can perform preliminary identification of landslide bodies when processing a large number of thermal images, their efficiency is low and it is difficult to effectively solve the noise and interference problems in the image. Especially in complex natural environments, the similarity between landslide bodies and other warming phenomena has a significant impact on the accuracy and reliability of the identification results.

[0004] During open-pit mining, slopes are often subjected to stress disturbances, which can lead to a decrease in slope stability, increase the risk of landslides, and directly threaten the safety of mine production and the lives and property of workers. Therefore, it is of great practical significance to conduct non-contact accurate identification of open-pit slope landslides. Summary of the invention

[0005] In view of the defects of the prior art, the present invention provides an open-air slope landslide body identification and early warning system and method based on infrared local warming effect.

[0006] In order to achieve the above invention object, the technical solution adopted by the present invention is as follows:

[0007] An open-air slope landslide identification and early warning system based on infrared local warming effect includes the following modules:

[0008] IR acquisition and temperature increase extraction module: used to collect infrared thermal images of the slope area through satellite far-field monitoring or drone near-field monitoring, and to clean the collected original infrared thermal images to delineate the slope area and remove interference information from non-slope areas;

[0009] IR local warming identification module: used to input the processed infrared thermal image into the infrared local warming intelligent identification model to identify the local warming area in the thermal image. The infrared local warming intelligent identification model is a U-Net image segmentation network based on deep learning;

[0010] Landslide comprehensive analysis and early warning module: used to identify the location and range of the landslide body according to the local warming area output by the infrared local warming intelligent identification model; conduct landslide risk assessment in combination with geological and meteorological data, and issue landslide early warning signals.

[0011] Furthermore, the IR local warming identification module adopts a U-Net deep learning network, which includes an encoder and a decoder. Jump connections are used to fuse the features of the encoder and the decoder to improve segmentation accuracy.

[0012] Furthermore, the IR acquisition temperature increase extraction module further comprises:

[0013] The infrared thermal image of the slope is collected by an infrared thermal imager, and the collected data includes the local temperature increase phenomenon of the rock during the landslide process.

[0014] The collected infrared thermal images are preprocessed to remove environmental interference signals and extract slope areas for effective monitoring.

[0015] Furthermore, the infrared local warming intelligent identification model is trained by the following steps:

[0016] Through the infrared thermal image data of marble and granite during the uniaxial loading fracture process, the thermal images with significant local temperature increase phenomenon are obtained;

[0017] The thermal image is expanded using data enhancement technology to generate a segmentation mask image containing the local temperature increase area as a training data set;

[0018] The U-Net deep learning network is used for model training, the Focal Loss loss function is used to solve the imbalance problem of positive and negative samples, the Adam optimizer is used for training, and the pre-trained weights of the VGG network are introduced to improve model performance.

[0019] Furthermore, the local temperature increase area in the infrared thermal image corresponds to the active area of ​​the landslide body. The activity of the landslide body is accompanied by soil friction and pressure changes, which leads to local temperature increase. The local temperature increase area is the area where the landslide body may occur.

[0020] Furthermore, the open-air slope landslide body identification and early warning system can monitor the slope area remotely and contactlessly, and issue landslide early warning signals in advance through real-time data processing and analysis.

[0021] The present invention also discloses an open-air slope landslide body intelligent identification method based on infrared local warming effect, comprising the following steps:

[0022] Step 1: Collect infrared thermal images of the slope area through satellite far-field monitoring or drone near-field monitoring, and clean the original infrared thermal images to remove interference signals in non-slope areas;

[0023] Step 2: Input the cleaned thermal image into the infrared local heating intelligent identification model to identify the local heating area and generate a heating area segmentation map;

[0024] Step 3: Based on the identified local warming areas, combined with geological and meteorological data, analyze and evaluate the potential risks of landslides and issue landslide warning signals.

[0025] Furthermore, the evaluation criteria of the infrared local warming intelligent identification model include:

[0026] Average pixel accuracy: used to evaluate the pixel accuracy of the model in each category. The calculation formula is as follows:

[0027]

[0028] Average IoU: It is used to evaluate the IoU of each category. It calculates the IoU of each category and averages it. The calculation formula is as follows:

[0029]

[0030] Among them, TP is the number of pixels correctly classified as the target category; TN is the number of pixels correctly classified as the non-target category; FP is the number of pixels incorrectly classified as the target category; FN is the number of pixels incorrectly classified as the non-target category; i is the specific category; N is the number of categories.

[0031] Furthermore, the local temperature increase area corresponds to the active area of ​​the landslide body, and the active area may indicate the occurrence of a landslide.

[0032] Furthermore, the landslide warning signal is issued based on the location, scope and risk assessment results of the landslide activity area.

[0033] The present invention also discloses a computer device, including a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the program, the above-mentioned intelligent identification method for open-air slope landslide bodies is implemented.

[0034] The present invention also discloses a computer-readable storage medium on which a computer program is stored. When the program is executed by a processor, the above-mentioned intelligent identification method for open-air slope landslide body is realized.

[0035] Compared with the prior art, the advantages of the present invention are:

[0036] 1. Based on the infrared local warming intelligent identification model, the present invention combines deep learning technology (such as U-Net network, VGG network and Adam algorithm) to achieve high-precision identification of open-air slope landslides. Experimental results show that the present invention has achieved relatively excellent results in recognition accuracy, with an mPA value of 94.37% and an mIoU value of 89.21%, which effectively improves the recognition ability of landslides.

[0037] 2. Non-contact monitoring of slopes by infrared thermal imagers avoids the safety risks caused by manual contact in traditional monitoring methods, and can achieve real-time monitoring of large areas, which is particularly suitable for monitoring slopes in complex and dangerous open-pit mines.

[0038] 3. The present invention can accurately identify the active area of ​​landslides by establishing an intelligent landslide identification and landslide warning system, and conduct comprehensive analysis and risk assessment based on geological, meteorological and other data, and issue landslide warning signals in a timely manner, thus providing strong support for the prevention and emergency response of landslide disasters.

[0039] 4. Compared with the traditional manual analysis method, the present invention significantly improves the efficiency and accuracy of landslide identification through automated image processing and intelligent models, reduces the impact of human factors on the judgment results, and can process a large amount of infrared thermal imaging data in a shorter time.

[0040] 5. The present invention can effectively cope with complex engineering site conditions, especially in an environment where the temperature increase of the landslide body coexists with the temperature increase of other non-landslide bodies (such as pedestrians, buildings, water bodies, forests, etc.). With the high efficiency of the intelligent identification system, it can effectively filter out interference information and reduce misjudgment and distortion.

[0041] 6. Through real-time monitoring and early warning of open-pit mine slopes, the present invention can effectively prevent the occurrence of landslide disasters, ensure safe production in mines, and protect the lives and property of mining workers, which has important practical application significance. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 11 is an example diagram of the expansion of the data set in an embodiment of the present invention; a is the original image, b is horizontally flipped, c is vertically flipped, and d is rotated 180°;

[0043] Figure 2 : is a processed thermal image of a rock destruction process according to an embodiment of the present invention; wherein a is the original image, and b is the segmentation mask;

[0044] Figure 3 This is a diagram of the intelligent identification system according to an embodiment of the present invention;

[0045] Figure 4 This is the local temperature increase recognition effect of the rock damage thermal image in the embodiment of the present invention; a is the original image, and b is the recognition effect image;

[0046] Figure 5 This is a diagram showing the change of Loss with Epoch in an embodiment of the present invention;

[0047] Figure 6 This is a diagram showing changes in mIoU and mPA with Epoch in an embodiment of the present invention;

[0048] Figure 7 It is a structural diagram of an open-air slope landslide body identification and early warning system according to an embodiment of the present invention;

[0049] Figure 8 This is a diagram showing the local temperature increase recognition effect of a thermal image of a slope according to an embodiment of the present invention. DETAILED DESCRIPTION

[0050] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below with reference to the accompanying drawings and examples.

[0051] 1. Intelligent identification system construction and testing

[0052] (1) Construction of intelligent recognition model dataset

[0053] Infrared monitoring tests of uniaxial loading fracture of marble and granite were carried out in the early stage, and it was found that the infrared thermal images of rocks changed from overall warming to local IRT strengthening, corresponding to the localization of fractures and stress regional concentration. For example, compressive shear stress concentration corresponds to IRT high temperature aggregation, tensile stress concentration corresponds to IRT low temperature differentiation, shear cracks are IRT high temperature strips, tensile cracks are IRT low temperature strips, covering type is IRT high temperature aggregation, and covering-uncovering type is related to the stress state of exposed rocks. During the deformation and destruction of rocks, there are a total of four types of local warming, namely regional high temperature aggregation, point-like high temperature aggregation, strip-like high temperature aggregation and high temperature dust radiation flow.

[0054] 104 thermal images with significant local warming phenomena were selected from the obtained thermal images to construct the data set. In order to improve the generalization ability of the intelligent recognition model, data enhancement techniques (such as rotation, flipping, scaling, etc.) were used to expand the data set size ( Figure 1 ), the total number of enhanced thermal image data sets reached 416. The infrared thermal image change map was segmented using the image segmentation model SAM (SegmentAnything Model) to generate a segmentation mask map containing the local temperature increase area ( Figure 2 ). Store the segmentation mask image in the same (Segmentation) folder to complete the data set creation.

[0055] (2) Model selection

[0056] like Figure 3 As shown in the figure, the deep learning network model selected by the intelligent recognition system is U-Net. U-Net is a deep learning network model designed for image segmentation tasks. Its structure is "U"-shaped and mainly consists of two parts: encoder and decoder. As a variant of the Fully Convolutional Network (FCN), U-Net was first proposed in image segmentation in the biomedical field and has now been extended to multiple application fields such as satellite imaging and autonomous driving. In the encoder part of U-Net, the resolution of the input image is gradually reduced through multi-layer convolution and pooling operations to extract high-level semantic features. The convolution layer is used to capture the high-level features of the image, and the pooling layer implements the downsampling process to reduce the size of the feature map. This process is repeated, allowing the network to gradually focus on more abstract information. The decoder part is the opposite of the encoder, and the low-resolution feature map is gradually upsampled to the resolution of the original image. Upsampling is performed through the deconvolution layer, and the features of the same resolution in the encoder are fused with the features in the decoder with the help of jump connections. This feature fusion strategy not only retains the detail information, but also enhances the accuracy of feature positioning, greatly improving the accuracy of the segmentation results.

[0057] (3) Intelligent identification system test

[0058] 1) Model training

[0059] The dataset was randomly divided into training set, validation set and test set in a ratio of 8:1:1 for model training, validation and testing. The initial learning rate of training was set to 1×10 -4, the training batch size is set to 20, and the total number of training rounds is set to 200 epochs. In order to understand the training progress of the model, the model is evaluated during the training process, and the evaluation frequency is once every 5 epochs. In order to enable the model to dynamically adjust the learning rate according to the historical gradient of each parameter, the Adam optimizer is used. In order to improve the performance of the model, the pre-trained weights of the VGG network are used during the model training process. In order to prevent the imbalance of positive and negative samples, the Focal Loss loss function is used.

[0060] 2) Evaluation parameters

[0061] Mean Pixel Accuracy (mPA) and mean Intersection over Union (mIoU) were selected as the evaluation criteria for the intelligent identification method of infrared local warming of rock damage.

[0062] Pixel Accuracy (PA) refers to the proportion of pixels correctly classified by the model in the segmentation task to the total pixels. Mean pixel accuracy (mPA) is the average value of the pixel accuracy of each category, which can provide a more comprehensive evaluation of the model performance. The calculation formula (Formula 1) is as follows:

[0063]

[0064] Among them, TP (True Positive) is the number of pixels correctly classified as the target category; TN (True Negative) is the number of pixels correctly classified as the non-target category; FP (False Positive) is the number of pixels incorrectly classified as the target category; FN (False Negative) is the number of pixels incorrectly classified as the non-target category; i is a specific category; N is the number of categories.

[0065] The intersection over union (IoU) indicates the degree of overlap between the predicted area and the true area. The mean intersection over union (mIoU) is calculated by averaging the IoU of each category to provide an overall performance evaluation. The calculation formula (Formula 2) is as follows:

[0066]

[0067] 3) Recognition effect

[0068] The trained model is used to identify the local temperature increase areas in the infrared thermal images of rock damage. Some of the recognition results are as follows: Figure 4 As shown in the figure, the model has a good recognition effect on the local warming area of ​​the thermal image.

[0069] Analyzing the training effect from the loss function value, the training set loss function value Train loss and the validation set loss function value Val loss both decrease as the number of epochs increases ( Figure 5 ), and finally flattens out and converges, which indicates the effectiveness of the training.

[0070] The experimental results are analyzed from the two indicators of mPA and mIoU. Both mPA and mIoU increase with the increase of epoch times and finally reach a stable value ( Figure 6 ), which shows the effectiveness of the model.

[0071] The various parameters of the experiment are shown in Table 1. Combined with the above analysis, it can be concluded that the intelligent identification system has excellent effect on the rock infrared local warming data set, with an mPA value of 94.37% and an mIoU value of 89.21%, which shows that the intelligent identification system can effectively identify the rock infrared local warming area.

[0072] Table 1 Experimental parameters

[0073]

[0074] 2. Landslide Identification and Early Warning System

[0075] like Figure 7 As shown in the figure, the landslide body identification and early warning system is mainly used for remote non-contact identification of slope landslide bodies and landslide early warning. The system includes three modules, namely, IR acquisition warming extraction layer, IR local warming identification layer, and landslide comprehensive analysis and early warning layer.

[0076] IR acquisition warming extraction layer: Use satellite far-field monitoring or drone near-field monitoring of slopes to collect infrared thermal images. The original infrared thermal images may contain a lot of environmental interference information (such as pedestrians, buildings, forests, water bodies, etc.). The heat changes in these non-slope areas will interfere with the identification of landslides. Therefore, it is necessary to clean the collected infrared thermal images and delineate the slope area.

[0077] IR local warming identification layer: The processed data is input into the infrared local warming intelligent identification model. The infrared local warming intelligent identification model is a deep learning image segmentation model built based on U-Net. The trained and optimized intelligent identification model can effectively identify the warming area in the infrared thermal image.

[0078] Landslide comprehensive analysis and early warning layer: The infrared local warming intelligent identification model outputs infrared thermal images of the identified local warming areas. The warming area usually corresponds to the active area of ​​the landslide body, because the landslide process is often accompanied by soil friction and pressure changes, resulting in local temperature increases. Through the infrared local warming intelligent identification model, the system can accurately identify the location and range of the landslide body. Integrate geological, meteorological and other data for risk assessment, and use the early warning system to determine whether there is a potential risk of landslides, and issue landslide warning signals in a timely manner.

[0079] The constructed landslide identification and early warning system is used to identify thermal images of slopes. Figure 8 As shown in the figure. It is not difficult to see from the recognition effect that the infrared warming intelligent identification system based on the U-Net deep learning network model built by the present invention has high recognition accuracy and successfully identifies the local warming area in the infrared thermal image. Combined with high-definition photos, after finding the slope, the slope landslide body can be successfully delineated.

[0080] In summary, by using the rock infrared local warming intelligent identification model based on the U-Net deep learning network, we carried out application research on remote non-contact identification of slope landslides, built a landslide identification and early warning system, and realized remote non-contact identification of landslides with good identification effect.

[0081] In another embodiment of the present invention, a terminal device is provided, which includes a processor and a memory, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is suitable for implementing one or more instructions, and is specifically suitable for loading and executing one or more instructions to implement the corresponding method flow or corresponding function; the processor described in the embodiment of the present invention can be used for the operation of the intelligent identification method of open-air slope landslide body, including the following steps:

[0082] Step 1: Collect infrared thermal images of the slope area through satellite far-field monitoring or drone near-field monitoring, and clean the original infrared thermal images to remove interference signals in non-slope areas;

[0083] Step 2: Input the cleaned thermal image into the infrared local heating intelligent identification model to identify the local heating area and generate a heating area segmentation map;

[0084] Step 3: Based on the identified local warming areas, combined with geological and meteorological data, analyze and evaluate the potential risks of landslides and issue landslide warning signals.

[0085] In another embodiment of the present invention, the present invention further provides a storage medium, specifically a computer-readable storage medium (Memory), which is a memory device in a terminal device for storing programs and data. It can be understood that the computer-readable storage medium here can include both the built-in storage medium in the terminal device and the extended storage medium supported by the terminal device. The computer-readable storage medium provides a storage space, which stores the operating system of the terminal. In addition, one or more instructions suitable for being loaded and executed by the processor are also stored in the storage space, and these instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory, or a non-volatile memory (non-volatile memory), such as at least one disk storage.

[0086] The processor may load and execute one or more instructions stored in a computer-readable storage medium to implement the corresponding steps of the method for intelligently identifying open-pit slope landslide bodies in the above embodiment; the processor may load and execute the following steps:

[0087] Step 1: Collect infrared thermal images of the slope area through satellite far-field monitoring or drone near-field monitoring, and clean the original infrared thermal images to remove interference signals in non-slope areas;

[0088] Step 2: Input the cleaned thermal image into the infrared local heating intelligent identification model to identify the local heating area and generate a heating area segmentation map;

[0089] Step 3: Based on the identified local warming areas, combined with geological and meteorological data, analyze and evaluate the potential risks of landslides and issue landslide warning signals.

[0090] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0091] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0092] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0093] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0094] Those skilled in the art will appreciate that the embodiments described herein are intended to help readers understand the implementation methods of the present invention, and should be understood that the protection scope of the present invention is not limited to such special statements and embodiments. Those skilled in the art can make various other specific variations and combinations that do not deviate from the essence of the present invention based on the technical revelations disclosed by the present invention, and these variations and combinations are still within the protection scope of the present invention.

Claims

1. An open-air slope landslide identification and early warning system based on infrared local warming effect, characterized in that: Includes the following modules: IR acquisition and temperature increase extraction module: used to collect infrared thermal images of the slope area through satellite far-field monitoring or drone near-field monitoring, and to clean the collected original infrared thermal images to delineate the slope area and remove interference information from non-slope areas; IR local warming identification module: used to input the processed infrared thermal image into the infrared local warming intelligent identification model to identify the local warming area in the thermal image. The infrared local warming intelligent identification model is a U-Net image segmentation network based on deep learning; Landslide comprehensive analysis and early warning module: used to identify the location and range of the landslide body according to the local warming area output by the infrared local warming intelligent identification model; conduct landslide risk assessment in combination with geological and meteorological data, and issue landslide early warning signals.

2. The open-air slope landslide body identification and early warning system according to claim 1 is characterized in that: The IR local warming identification module adopts the U-Net deep learning network, which includes an encoder and a decoder. Jump connections are used to fuse the features of the encoder and the decoder to improve segmentation accuracy.

3. The open-air slope landslide body identification and early warning system according to claim 1 is characterized in that: The IR acquisition temperature increase extraction module further comprises: The infrared thermal image of the slope is collected by infrared thermal imager, and the collected data includes the local temperature increase of the rock during the landslide process; The collected infrared thermal images are preprocessed to remove environmental interference signals and extract slope areas for effective monitoring.

4. The open-air slope landslide body identification and early warning system according to claim 1 is characterized in that: The infrared local warming intelligent identification model is trained by the following steps: Through the infrared thermal image data of marble and granite during the uniaxial loading fracture process, the thermal images with significant local temperature increase phenomenon are obtained; The thermal image is expanded using data enhancement technology to generate a segmentation mask image containing the local temperature increase area as a training data set; The U-Net deep learning network is used for model training, the Focal Loss loss function is used to solve the imbalance problem of positive and negative samples, the Adam optimizer is used for training, and the pre-trained weights of the VGG network are introduced to improve model performance.

5. The open-air slope landslide body identification and early warning system according to claim 1 is characterized in that: The local temperature increase area in the infrared thermal image corresponds to the active area of ​​the landslide body. The activity of the landslide body is accompanied by soil friction and pressure changes, which leads to local temperature increase. The local temperature increase area is the area where the landslide body may occur.

6. The open-air slope landslide body identification and early warning system according to claim 1 is characterized in that: The open-air slope landslide body identification and early warning system can monitor the slope area in a remote non-contact manner, and issue a landslide early warning signal in advance through real-time data processing and analysis.

7. An intelligent identification method for open-air slope landslide based on infrared local warming effect, characterized by: The method is implemented on the basis of the open-air slope landslide body identification and early warning system according to any one of claims 1 to 6, and comprises the following steps: Step 1: Collect infrared thermal images of the slope area through satellite far-field monitoring or drone near-field monitoring, and clean the original infrared thermal images to remove interference signals in non-slope areas; Step 2: Input the cleaned thermal image into the infrared local heating intelligent identification model to identify the local heating area and generate a heating area segmentation map; Step 3: Based on the identified local warming areas, combined with geological and meteorological data, analyze and evaluate the potential risks of landslides and issue landslide warning signals.

8. The method for intelligent identification of open-air slope landslide according to claim 7 is characterized in that: The evaluation criteria of the infrared local warming intelligent identification model include: Average pixel accuracy: used to evaluate the pixel accuracy of the model in each category. The calculation formula is as follows: Average IoU: It is used to evaluate the IoU of each category. It calculates the IoU of each category and averages it. The calculation formula is as follows: Among them, TP is the number of pixels correctly classified as the target category; TN is the number of pixels correctly classified as the non-target category; FP is the number of pixels incorrectly classified as the target category; FN is the number of pixels incorrectly classified as the non-target category; i is the specific category; N is the number of categories.

9. A computer device, characterized in that: The method comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the program, the method for intelligently identifying open-air slope landslide bodies as described in claim 7 is implemented.

10. A computer-readable storage medium, characterized in that: A computer program is stored thereon, and when the program is executed by a processor, the method for intelligently identifying open-air slope landslide bodies as described in claim 7 is implemented.

Images