30 results about "Landslide detection" patented technology

The invention discloses a lightweight-based satellite landslide image intelligent detection method, device and equipment and a medium, and relates to the technical field of disaster detection, and the method comprises the steps: obtaining a whole-scene optical satellite image containing a landslide and a non-landslide region and landform auxiliary data; a dynamic segmentation strategy is adopted to carry out differential segmentation and standardized preprocessing on an image based on topographic data, and a standardized image is obtained. A target landslide image is screened through a double-layer machine learning model, the target image is input into an improved lightweight convolutional neural network for processing, an initial detection result is obtained, finally edge optimization and coordinate calibration are performed on the result, and an accurate landslide area detection result is output. The identification precision of the landslide image is effectively improved through dynamic segmentation and double-layer screening, the improved lightweight convolutional neural network realizes efficient detection in a low-resource environment, and the accuracy of a landslide detection result is further improved.

Disclosed are systems and methods that employ elastic fiber optic, time-of-flight sensors for long distance landslide monitoring with sub-mm precision. The time-of-flight (ToF) sensor is integrated in conjunction with a single-photon avalanche diode (SPAD). By coupling both the emission source and the detector with a stretchable optical fiber, our inventive systems and methods continuously monitor the length of the stretchable optical fiber by measuring a traveling time of an optical pulse traversing the stretchable optical fiber. A significant, detectable change in the length of the stretchable optical fiber – indicative of ground movement or deformation, triggers an alarm, providing an early warning for potential landslides. As such, systems and methods according to aspects of the present disclosure provide a reliable, sensitive, precise, cost-effective, real-time solution for landslide detection and monitoring – a problem that has plagued the art.

The invention discloses a landslide detection deep learning method fusing geometry, context and frequency domain multi-source features, and belongs to the technical field of geological disaster recognition and landslide remote sensing information processing. The problem that a traditional landslide detection method in the prior art is low in recognition precision is solved. The method comprises the following steps: constructing a landslide detection model based on a GCF-Net architecture, and inputting a to-be-detected image into the landslide detection model; capturing feature information of different scales in the to-be-detected image through a ghost pyramid feature fusion module; an output characteristic pattern is input to a decoder through a GCFmodule; and an EdgeECI module in the decoder is used to process the output feature map of the GCFI module, and a final output feature map after fusion is obtained. The landslide detection accuracy is effectively improved, and the method can be applied to geological disaster monitoring and early warning, mountain traffic and infrastructure safety, water conservancy and energy engineering operation monitoring, remote sensing big data processing platforms and other scenes.

The present application relates to the technical field of geological detection, and particularly relates to a landslide detection device, which comprises a shell, a friction induction surface on the inner wall of the shell, an electro-induction turntable connected to the shell in a relative rotating manner, an electro-induction boss on the circumferential side of the electro-induction turntable, a driving assembly comprising a driving member arranged outside the shell and a transmission member connected with the driving member and extending into the shell for driving the electro-induction turntable to rotate, and a circuit system arranged in the shell and electrically connected with the friction induction surface, wherein the circuit system is configured to capture an electric signal generated by the friction between the electro-induction boss and the friction induction surface when the driving member drives the transmission member to drive the electro-induction turntable to rotate under an external force, and transmit the electric signal to an external monitoring device. The present application converts the ground movement during a landslide into the rotation of the electro-induction turntable and the friction between the electro-induction boss and the friction induction surface to generate an electric signal, and realizes the detection of the landslide by capturing the electric signal.

The application discloses a SAR image earthquake landslide boundary detection method and system based on deep learning, and the method comprises the following steps: inputting two SAR images before and after a landslide to generate a normalized mean ratio image, inputting the normalized mean ratio image into a pre-trained landslide detection network model based on deep learning after upsampling, extracting a landslide, extracting a boundary, and obtaining a landslide edge detection result; the landslide detection network model comprises an encoder and a decoder, the encoder comprises a backbone network and a coordinate attention empty space pyramid module, and the decoder comprises an attention adjustment module and a multi-scale feature fusion resampling module.The application can realize rapid, high-precision and automatic detection of large-scale earthquake landslide boundaries, and can realize detection of different earthquake types and different sizes of earthquake landslides.

The invention discloses a landslide displacement measurement method and system, and relates to the technical field of landslide detection, and the method comprises the steps: burying a detection inductor on a landslide body, connecting the two ends of the detection inductor to the top of the landslide body through insulated wires, connecting a voltmeter, collecting monitoring data, and carrying out the preprocessing; measuring the magnetic field intensity and the magnetic induction line direction on the landslide mass, calculating the included angle between the magnetic induction line and the detection inductor, constructing a correction model, calculating correction parameters, and obtaining displacement parameters; the monitoring voltage of the voltmeter is recorded, the actual displacement of the detection inductor in the magnetic induction line is calculated, and the landslide displacement is calculated and judged by combining displacement parameters. According to the invention, the detection is accurate, the external environment interference is small, the applied detection equipment is less, the possibility of landslide slippage can be accurately analyzed, the detection of landslide displacement at different depths can be met, and the possibility of damage to the detection equipment is reduced.

The invention provides a landslide area extraction method and device, electronic equipment and a storage medium, and belongs to the technical field of image data processing. The method comprises the following steps: respectively inputting a first remote sensing image of a time phase before occurrence of a landslide, a second remote sensing image of a time phase after occurrence of the landslide and a slope map into a first feature extraction network, a second feature extraction network and a third feature extraction network of a landslide extraction depth model, a first feature map, a second feature map and a third feature map which are correspondingly output are obtained; inputting the first feature map, the second feature map and the third feature map into a feature fusion module of a landslide extraction depth model to obtain a fused feature map output by the feature fusion module; and inputting the fused feature map into a decoder of the landslide extraction depth model to obtain a landslide extraction result output by the decoder. According to the method, the accuracy, robustness and efficiency of landslide automatic detection and mapping are improved, the slope map is introduced as a supplement of topographic information, and the identification capability of the model for the potential landslide area is enhanced.

The present application relates to the technical field of remote sensing image, and particularly relates to a highland mountainous area landslide detection method based on an improved YOLOv11 deep learning model, comprising: obtaining remote sensing images of a target area and performing pretreatment; constructing a YOLOv11 target detection model, and introducing a spatial adaptive feature modulation module and a convolution channel mixer into the model, the former is used for adaptively recalibrating the spatial dimension of a feature map to enhance the landslide area feature representation, and the latter is used for extracting local context information; using an improved point distance-based intersection over union loss function as a positioning loss, and training the model in combination with a multi-scale training strategy; inputting the image to be detected into the trained model, and outputting a landslide detection result. The cascade design of long-range feature modeling and local feature extraction enhances the feature extraction capability of the model for landslide targets, the loss function improves the boundary box regression accuracy, and effectively improves the accuracy and robustness of landslide detection in the complex background of highland mountainous areas.

This invention discloses a method and device for geological disaster prevention and control of power transmission and transformation tower foundations, belonging to the field of landslide detection technology. The method involves: real-time acquisition of optical remote sensing images, radar remote sensing images, microwave remote sensing images, and thermal infrared remote sensing images; overlaying optical and radar remote sensing images to generate slope topographic spatial distribution data; analyzing vegetation growth status in the optical remote sensing images to obtain vegetation spatial distribution data; overlaying the slope topographic spatial distribution data and vegetation spatial distribution data to obtain topographic-vegetation coupled spatial distribution data, thereby determining the landslide sensitivity level of each preset grid unit and generating landslide sensitivity level spatial distribution data; and retrieving water storage spatial distribution data based on microwave and thermal infrared remote sensing images to assess the stability of the tower foundation area where the power transmission and transformation equipment is located, obtaining landslide level distribution data to determine control measures. This invention solves the problem of low accuracy in landslide detection in existing technologies.

The application discloses a remote sensing image landslide detection method and system, and belongs to the technical field of landslide detection. The method adopts a data enhancement method to expand a landslide data set, and solves the problem of a small number of landslide samples in a training set. DarkNet53 is used to replace a commonly used backbone feature extraction network of FasterR-CNN, so that the problem of gradient disappearance caused by a too deep network is avoided, and the detection precision is improved.

The invention relates to the technical field of landslide detection, in particular to a high-position hidden landslide blind area intelligent monitoring system based on the Internet of Things. Comprising the following steps: S1, accessing multi-source heterogeneous time series data from an Internet of Things monitoring network arranged on a landslide mass, the multi-source heterogeneous time series data comprising deep displacement data, ground surface displacement data, pore water pressure data and rainfall data; s2, preset sensitive characteristic parameters are extracted from the multi-source heterogeneous time sequence data respectively, all the sensitive characteristic parameters are aligned to the same time reference in a unified mode, a multi-dimensional characteristic vector is obtained, and the sensitive characteristic parameters comprise displacement acceleration and osmotic pressure change rate; s3, performing correlation analysis according to the change trend of the displacement acceleration and the osmotic pressure change rate, and calculating a current abnormal index according to a correlation analysis result; s4, comparing the abnormal index with a preset threshold value defined based on a multi-parameter joint judgment rule; and when the combined judgment rule is met, judging that the landslide mass enters an abnormal state, and generating a corresponding early warning signal.

The application provides a kind of high-resolution image landslide detection method of combined spectrum, digital elevation model and auxiliary feature dual-branch network, including obtaining the high-resolution remote sensing image to be detected, digital elevation model and landslide true label;Extract auxiliary features, mine multiple characteristic factors from them and merge all images in channel;The image after merging is preprocessed, and landslide sample is generated;Different data enhancement means are used to increase the number and diversity of landslide data set;Residual and channel attention are added to the network model based on U-Net+++ to improve the network model;The image sample pair is respectively input into the improved U-Net+++ branch network, and the prediction result is obtained;The consistency of sample pair is mined and the model is optimized using the loss function;Based on the test sample, the model accuracy is evaluated and the landslide detection result map is output.The application provides a novel and effective landslide detection method, which can fully mine the invariance features of high-resolution landslide images and effectively improve the recognition accuracy of landslide.

The invention discloses mine landslide multi-view sound wave detection equipment and a mine landslide multi-view sound wave detection system, and relates to the technical field of mine landslide detection equipment. The mine landslide multi-view sound wave detection equipment comprises an outer sleeve, spiral plates are fixedly installed on the outer side of the outer sleeve, the ends, away from the outer sleeve, of the spiral plates are bent upwards, the spiral plates are evenly installed on the outer side of the outer sleeve in the axial direction, and separation rings are fixedly installed on the outer side of the outer sleeve. In cooperation with separation of a separation ring, when hole gaps are installed and filled and hole sealing treatment is carried out, backfilled soil is blocked in a segmented mode through the separation ring, meanwhile, the side, away from the outer sleeve, of the spiral plate is bent upwards, part of soil backfilled with the soil is fixed, and connection is formed between the outer sleeve and the inner wall of the hole through the gap; and the phenomenon that a gap exists between the outer sleeve and the inner wall of the hole, so that fluctuation is blocked due to the gap when the fluctuation is accepted, the acceptance of seismic source fluctuation is influenced, and detection data is influenced is prevented.

The invention provides a self-supervision landslide change detection method and system based on NDVI driving and feature interaction. The method comprises the following steps: acquiring a dual-temporal remote sensing image of a disaster area to generate pre-disaster and post-disaster NDVI images; preprocessing the dual-time-phase remote sensing image to generate a dual-time-phase three-channel optical image; the double-time-phase three-channel optical image is input into an unsupervised landslide detection model, the model is configured to extract double-time-phase features and calculate feature differences, and a landslide prediction probability is output; a landslide pseudo label is generated based on the difference value of the NDVI images before and after the disaster, and the landslide prediction probability is constrained by using a cross entropy loss function; aligning the features of the same category extracted from different images based on a category perception feature alignment loss function; iteratively updating the unsupervised landslide detection model in combination with the cross entropy loss function and the category perception feature alignment loss function to obtain an optimized model; and carrying out landslide detection on the dual-time-phase remote sensing data of the target area by using the optimized model, and outputting a result.

The invention relates to the technical field of landslide detection, in particular to an unmanned aerial vehicle image landslide post-disaster evaluation method based on FYOLOv11, and mainly solves the technical problems that an existing landslide recognition method based on the deep learning technology is high in model complexity, low in reasoning speed and difficult to deploy on edge equipment. The method comprises the following steps: S1, acquiring a data set; s2, constructing a YOLOv11 model, and replacing an original BackBone network with a FasterNet network so as to form a FYOLOv11 model; s3, based on the data set, training an FYOLOv11 model to obtain a target model; and S4, taking the unmanned aerial vehicle image as input, and outputting a landslide identification result through the target model. According to the method, the parameter quantity and the calculation complexity can be remarkably reduced while the feature expression capability is kept, the advantages of monitoring precision, reasoning speed and light weight are considered, and the applicability of the model on edge equipment is improved.

The invention relates to the technical field of debris flow landslide detection, and discloses a railway debris flow landslide detection device which comprises an arc-shaped rain baffle, a storage battery box and a debris flow infrasound monitor, the storage battery box and the debris flow infrasound monitor are installed at the bottom of the arc-shaped rain baffle, and a square plate is fixed to the bottom of the storage battery box; the four mounting plates are fixed to the bottom of the square plate and distributed in an annular array mode, a telescopic frame is arranged below the bottoms of the mounting plates, a connecting block is fixed to the middle of one end of the telescopic frame, and a fixing piece is installed on the connecting block; under the action of the designed swing unit, the synchronous unit and the power unit, natural wind power is used as driving force to drive the swing unit to swing back and forth, so that the detection position of the detection probe is automatically changed, the flexibility of the detection probe is improved, the detection range is expanded, and the detection efficiency is improved. The debris flow monitoring device can monitor debris flow in a large range, and is more energy-saving and efficient.

The application provides a kind of optical remote sensing image landslide disaster automatic identification method and system, belong to image processing technical field.The method includes: multi-view feature extraction and fusion are carried out to input image, and robust feature representation is obtained;Based on this, initial landslide detection frame and semantic segmentation map are synchronously generated by integrated model;Further, an iterative optimization cycle containing segmentation quality evaluation, detection frame correction and segmentation map optimization guidance is executed, so that detection and segmentation tasks dynamically interact and enhance in inference process until the convergence condition is met;Finally, high-precision landslide positioning frame and pixel-level segmentation mask are output.The system is used to execute the method.The application adopts the above-mentioned optical remote sensing image landslide disaster automatic identification method and system, effectively solves the core contradiction that landslide positioning is not accurate and boundary recognition is rough and difficult to consider in the prior art, improves the recognition accuracy, robustness and regional adaptive ability of the model in complex scenes.

The present application relates to landslide detection technical field, especially in landslide detection device and method, the device includes the detection pipe body buried in the monitoring area underground or surface and laser receiving calibration mechanism, the inside of detection pipe body is provided with plumb line, the upper end of plumb line is fixed on the stable support of detection pipe body upper end, and the lower end is hung with counterweight;The bottom of detection pipe body is located below the counterweight and is provided with angle disc, the annular photosensitive scribe line array is arranged on the angle disc, and is used for real-time recording plumb line swing direction and amplitude change.The landslide detection device provided by the present application can instantly generate corresponding photosensitive signal difference when the plumb line is slightly offset, and the signal processing module can accurately calculate the angle deviation of the plumb line relative to the vertical direction by analyzing the amplitude and timing change of the photosensitive unit output, so as to realize high-resolution monitoring of stratum microinclination and local deformation.

The application relates to a landslide detection method and device, computer equipment, a computer readable storage medium and a computer program product. The method comprises the following steps: acquiring historical landslide data of a target area, performing image correction and spatial registration processing to obtain processed historical landslide data, and constructing a landslide sample set; superimposing multi-source factor data and landslide label data in the channel dimension to obtain a multi-channel stereo image block, performing slice processing to obtain a training sample set and a test sample set; training a preset model based on the training sample set and the test sample set, obtaining a trained preset model under the condition that a preset training condition is met; calculating the performance index of the trained preset model, and acquiring real-time landslide data of the target area under the condition that the performance index meets a preset performance condition, inputting the real-time landslide data into the trained preset model, and obtaining a landslide detection result of the target area. The method can improve the accuracy of landslide detection.

The application discloses a video landslide monitoring and positioning method based on a geographic position positioning algorithm, and comprises the following steps: obtaining a landslide detection model by training a target detection network; obtaining a landslide positioning model by training a landslide positioning network; selecting a to-be-measured area, acquiring remote sensing images of the to-be-measured area and real-time to-be-measured photos; sequentially inputting the to-be-measured photos into the landslide detection model, outputting position coordinates of a landslide area in the to-be-measured photos when the landslide area is identified in the to-be-measured photos, and cutting out to-be-measured image slices from the to-be-measured photos; and inputting the to-be-measured image slices and the remote sensing images of the to-be-measured area into the landslide positioning model to obtain longitude and latitude coordinates of the landslide area. The application can realize real-time monitoring of the to-be-measured area and automatically output accurate positioning of the landslide area. The application is not affected by environment and weather, has simple requirements on equipment, is accurate in positioning, and has a small error range.

The invention provides an unmanned aerial vehicle landslide identification method based on wavelet domain feature separation and semantic decoupling, and relates to the technical field of landslide identification. Firstly, an unmanned aerial vehicle is used for carrying an optical sensor to carry out aerial survey on a landslide area to obtain a high-resolution remote sensing image; landslide area labeling is carried out on the image, and an unmanned aerial vehicle landslide semantic segmentation data set including a training set, a verification set and a test set is formed; then preprocessing the high-resolution remote sensing image; constructing a semantic segmentation model based on multi-scale feature enhancement and spatial attention fusion, and training the semantic model; performing quantitative and qualitative evaluation on the trained model on the verification set and the test set; deploying the optimized model to an unmanned aerial vehicle-mounted computing platform; landslide detection is carried out, and a landslide segmentation result and a boundary diagram are generated; and outputting a segmentation result to realize unmanned aerial vehicle landslide identification.

The invention relates to the technical field of landslide detection, in particular to a landslide detection device and method.The device comprises a detection pipe body and a laser receiving calibration mechanism, the detection pipe body is buried underground or on the earth surface of a monitoring area, a plumb line is arranged in the detection pipe body, and the upper end of the plumb line is fixed to a stable support at the upper end of the detection pipe body; a counterweight body is hung at the lower end; an angle scale is arranged at the bottom of the detection pipe body and located below the balance weight body, and an annular photosensitive scribed line array is distributed on the angle scale and used for recording the swing direction and amplitude change of a plumb line in real time. According to the landslide detection device provided by the invention, the annular photosensitive scribed line array can instantly generate corresponding photosensitive signal difference when the plumb line generates extremely small deviation, and the signal processing module can accurately solve the angular deviation of the plumb line relative to the vertical direction by analyzing the amplitude and time sequence change output by the photosensitive unit; and high-resolution monitoring of micro-inclination and local deformation of the stratum is realized.

The invention discloses a landslide detection method and device for an aerial image of an unmanned aerial vehicle, and the method comprises the steps: collecting various landslide data, and carrying out the format conversion, data enhancement and other first data processing operations, thereby improving the quality of data needed by model training, and improving the generalization capability of a model; then, based on a YOLOv11 model, constructing a landslide detection model of a layered architecture, and training by using the processed data to obtain a target landslide detection model, so that targeted training of the target landslide detection model is realized; moreover, before landslide detection is carried out on the image based on the target landslide detection model, second data processing operations such as real-time cloud concentration detection and adaptive defogging can be carried out on the to-be-detected image, and through the pre-defogging processing, the interference of cloud on the image is effectively removed, and the image quality is improved. Finally, efficient and accurate landslide detection is realized by the target model, and the accuracy and reliability of a landslide detection result output based on the target landslide detection model are improved.

The invention relates to the technical field of geological disaster monitoring and remote sensing intelligent identification, and particularly discloses a physical prior information constraint-based landslide intelligent extraction method, which comprises the following steps of: collecting and preprocessing environmental factor data, and constructing a multi-source landslide factor data set; foS is obtained through a TISSA model, and the FoS is fused with an optical image to construct a multi-source landslide detection data set; and constructing a PhysSeg-Dual model: on the basis of a YOLOv8 instance segmentation framework, constructing an independent double-branch parallel structure of an optical feature branch and a physical constraint feature branch, and introducing a multi-frequency frequency channel attention module in a feature fusion stage. The method effectively improves the robustness and precision of landslide recognition in a complex landform environment, has remarkable theoretical innovation and application and popularization values, and can be widely applied to the fields of geological disaster monitoring, post-disaster evaluation, land safety protection and the like.

This invention relates to the field of landslide detection equipment technology, and discloses an anti-collision type landslide monitoring device for geotechnical engineering. It includes a load-bearing plate with fixing mechanisms at each of the four bottom corners. An outer housing cylinder is installed at the top center of the load-bearing plate, and a solar panel is installed on top of the outer housing cylinder. Buffer foam is installed inside the outer housing cylinder, and monitoring equipment is installed inside the buffer foam. A protective cover is installed on top of the outer housing cylinder, and two buffer components are symmetrically arranged on both sides of the outer housing cylinder. This invention effectively dissipates frontal and lateral impact forces by incorporating multiple sliding and spring components; its bottom fixing mechanism utilizes gravity wedge action to achieve self-locking anchoring, enhancing stability; simultaneously, the modular design of the impact plate allows for quick replacement, and the top suspended protective cover uses spring buffering to ensure the safety of the solar panel.

The application provides a satellite landslide image detection method based on a multi-level Transformer network, and belongs to the field of image anomaly detection. In view of the fact that the existing landslide detection is only performed by using a convolutional neural network to extract and reconstruct image features, the application adopts a pre-training feature extraction network and an unsupervised multi-level Transformer auto-encoder to perform landslide detection. First, multi-scale features are extracted from the pre-training network, and then the image features are reconstructed by using a self-encoding Transformer network with a U shape. Through the fusion and reconstruction of the features by the multi-level Transformer encoder and decoder, coarse-grained low-level pixel features and fine-grained global semantic features can be obtained at the same time. The proposed method is applied to a satellite captured Bijie landslide data set, and experimental results show that, compared with an original CNN model, the application can improve detection accuracy and effectively distinguish landslide and non-landslide image data.

The utility model provides a kind of track area foreign matter invasion monitoring early warning system, including the monitoring point of several that along track along line interval arrangement;The monitoring point is provided with: vibration sensing module, for collecting the vibration signal in target track section;Image acquisition module, for collecting the image signal in target track section;Data processing module, the signal output end of vibration sensing module, image acquisition module and the signal input end electrical connection of data processing module;In the track section of existing side slope, the monitoring point further includes: setting in the landslide detection device of track area corresponding side slope, the signal output end of the landslide detection device and data processing module corresponding signal input end connection;The data processing module and dispatching center communication connection, mutual communication.This track area foreign matter invasion monitoring early warning system can monitor whether track area has foreign matter falling or invasion, also monitor landslide condition, early warning track area possible facing invasion.

The application discloses a landslide detection method based on adaptive fusion of optical satellite remote sensing and DEM data, and takes optical satellite remote sensing data and DEM data as core inputs to construct a heterogeneous double-branch high-resolution network: the optical branch adopts HRNet-W48 to extract spectral texture features, and the DEM branch adopts a lightweight HRNet-W18 to extract terrain geometric features. A terrain perception gating fusion module is designed, a pixel-level weight mask is dynamically generated based on the terrain features reflected by the DEM data through the gating network, and the adaptive weighted fusion of optical and terrain features is realized in the form of residual error; a CE+BCL combined loss function is proposed, and the boundary contrast loss is used to strengthen the supervision of the landslide edge area. The application effectively solves the problems of rigid fusion strategy, loss of spatial details and fuzzy boundary in the traditional method, and significantly improves the accuracy, boundary clarity and complex scene adaptability of landslide recognition.

The invention discloses an efficient landslide remote sensing detection method and system based on Transform, a storage medium and equipment, and belongs to the field of landslide detection.The method comprises the steps that S1, a landslide remote sensing image to be detected is obtained and preprocessed; s2, inputting the preprocessed image into an encoder, and extracting feature maps of different scales through overlapping block embedding and multi-stage transformation; s3, performing semantic-guided feature fusion on the multi-scale feature map output by the encoder to obtain fused features; and S4, inputting the fused features into a decoder, recovering the fused features to the size of an input image through lightweight convolution and up-sampling operation, outputting the probability that each pixel point is a landslide, and generating a final landslide detection binary image or probability graph. According to the method, overlapped convolution window processing, an approximate attention mechanism, implicit spatial position information modeling and semantic alignment and boundary reconstruction are introduced, so that the reasoning speed is remarkably increased while high detection precision is kept, and the method has good engineering application value and migration generalization ability.

The application discloses a kind of light-based satellite landslide image intelligent detection method, device, equipment and medium, it is related to disaster detection technical field, including: by obtaining the whole scene optical satellite image and topographic auxiliary data containing landslide and non-landslide area, based on topographic data, using dynamic segmentation strategy to image is differentially segmented and standardized preprocessing, obtain standardization image. Through double-layer machine learning model screening target landslide image, the target image is input to improved light convolutional neural network and is processed, and initial detection result is obtained, finally the result is edge optimized and coordinate calibration, and the accurate landslide area detection result is output. Through dynamic segmentation and double-layer screening, the recognition accuracy of landslide image is effectively improved, and the improved light convolutional neural network realizes efficient detection in a low-resource environment, further improves the accuracy of landslide detection result.