34 results about "Change detection algorithms" patented technology

The invention provides an automatic action trajectory labeling system and method based on a visual language model, and relates to the field of visual technical models. The automatic action trajectory labeling system based on the visual language model comprises a key frame extraction module, a VLM semantic labeling module, a historical labeling database, a diffusion model optimization module and an reflection correction module, and the key frame extraction module extracts a key frame sequence from an input video through optical flow analysis and a TSN network scene change detection algorithm; and the VLM semantic annotation module is used for outputting a candidate segmentation point set containing action starting and ending points, the VLM semantic annotation module adopts a multi-modal VLM, takes a key frame image and a context text instruction as input, and generates a preliminary semantic tag and a corresponding vision-language embedding vector. According to the method, multiple bottlenecks of the existing action track labeling technology are solved, the labeling precision is remarkably improved, and the target of time boundary optimization is achieved.

The present invention discloses an optical remote sensing image change detection method based on deep learning, belonging to the field of remote sensing image change detection. First, optical remote sensing image data is acquired and preprocessed; then, a two-step progressive structure is adopted to combine the Resnet18 network and the Vision Transformer model to extract the features of the images before and after the change; next, the extracted features are input into the channel attention module to obtain features with different attention sizes for different channels; then, the features before and after the change are sent into the difference information extraction module to obtain the difference feature map of the two remote sensing images; finally, the difference feature map is sent into the classifier to obtain the final change detection binary image, and it is compared with the change reference map. The present invention effectively optimizes and improves the problems existing in the existing change detection methods, such as low detection accuracy, large probability of false detection and missed detection, and poor universality of the change detection algorithm on different data sets.

The invention discloses an intelligent cultivated land outflow pattern spot extraction method. The method comprises the steps of data collection and preprocessing; samples are rapidly produced based on stock data and a manual collection mode, and sample metadata and file data are stored in a database mode and a file mode respectively; selecting a proper semantic segmentation and change detection algorithm to carry out adaptive transformation and combination, introducing a fully trained general segmentation and general change detection pre-training model, and training an iterative optimization network to obtain an uncultivated land identification and cultivated land change detection model with optimal precision; suspected cultivated land outflow pattern spots are extracted based on an optimal precision model; performing batch regularization processing on the suspected cultivated land outflow pattern spots to obtain suspected cultivated land outflow preliminary result pattern spots; and checking the correctness of the boundary and the attribute of the preliminary result pattern spot to obtain a suspected cultivated land outflow result pattern spot. According to the invention, the farmland outflow condition of the mountain area can be monitored, and the change pattern spots and type information can be extracted and updated quickly, efficiently and accurately.

The invention discloses a remote sensing image change detection algorithm based on multi-dimensional feature collaborative modeling. The method comprises the following steps: preprocessing and dividing image blocks of a remote sensing image change detection common reference data set; constructing a remote sensing image change detection model based on multi-dimensional feature collaborative modeling, inputting the output from a feature extraction stage into a double-branch feature difference modeling module by a trunk branch, and performing difference modeling on double-time-phase features of each layer to form multi-scale difference features; then inputting into a boundary protection feature calibration module to strengthen the edge, and inputting the strengthened features into a decoder to generate a detection result; the auxiliary branch inputs the data into a style difference feature alignment module; and training the model and then testing the test set. Compared with an existing change detection method FC-Siam-Conv, the four evaluation indexes including the accuracy rate, the recall rate, the F1 score and the intersection-union ratio are improved by 3.72%, 6.93%, 5.18% and 8.38% respectively; compared with the existing change detection method, the BIT is respectively improved by 1.63%, 1.11%, 1.38% and 2.32%.

A computer vision system for automatic identification, tracking and management of inventory and / or assets, wherein: the computer vision system is programmed with predetermined and configurable image data conditions that identify and capture certain types of movement; the computer vision system is programmed to constantly capture and process image data from one or more sets of sensors to detect the image data conditions; the computer vision system is programmed to trigger the process of identifying all objects and extracting label information from objects present in a field of view of the system if the image data conditions are fulfilled; the computer vision system processes the image data with one or more object detection algorithms and label detection algorithms and generate correspondences between objects and labels identified; and the computer vision system running change detection algorithms and semantic analysis to detect if any object was moved, removed, occluded and if any new objects were added in to the system.

The invention discloses a forest weak disturbance area identification method and device, equipment, a medium and a product, and relates to the technical field of forest detection, and the method comprises the steps: employing a change detection algorithm of a long-time sequence image to construct a prediction model according to a normalized red wave and short wave infrared distance index and a tasseled cap transformation humidity index; constructing a training set according to the two prediction models; each sample in the training set comprises input data and label data, the input data comprises waveform coefficients and environment data of the two prediction models, and the label data comprises a forest type and a non-forest type; the forest type comprises existence of forest weak disturbance and absence of forest weak disturbance; the detection threshold value for distinguishing the existence of the forest weak disturbance and the absence of the forest weak disturbance is obtained by performing adaptive calculation on a prediction model change probability threshold value by utilizing a gradient lifting decision tree algorithm; and training the classification model by adopting the training set to obtain a forest weak disturbance recognition model. The forest weak disturbance detection accuracy is improved.

A river and lake shoreline monthly monitoring remote sensing image automatic interpretation and result management system comprises a data resource management and access module, a remote sensing image intelligent preprocessing module, an intelligent interpretation and change identification module, a man-machine collaborative interaction and correction module and a database and report generation module. The data resource management and access module is used for accessing image data, the remote sensing image intelligent preprocessing module is used for data preprocessing, the intelligent interpretation and change recognition module is used for element automatic extraction and change discovery, and the man-machine collaborative interaction and correction module is used for man-machine interaction and result correction. And the database and report generation module is used for data storage and monitoring result generation. The invention provides an improved self-adaptive gating network sub-pixel shoreline interpretation algorithm for automatic interpretation of remote sensing images, and provides an improved cross-temporal attention and uncertainty collaborative change detection algorithm for monthly dynamic analysis of river and lake shorelines. A better scheme is provided for a river and lake shoreline monthly monitoring remote sensing image automatic interpretation and result management system.

The invention discloses an unmanned aerial vehicle intelligent ecological checking system with multi-source data fusion, and relates to the technical field of ecological environment protection and monitoring. The system comprises an unmanned aerial vehicle platform, a multi-source data acquisition module, a data processing and analysis module, an intelligent decision module, a communication and data transmission module and a ground control and management platform. According to the invention, by integrating the high-resolution optical camera, the multispectral camera and the thermal imager multi-source data acquisition module, multi-angle and all-directional data acquisition is realized, cooperative work of the devices on the unmanned aerial vehicle platform is combined with an advanced data processing and analysis algorithm, an earth surface change area can be accurately detected, and the change degree and property can be evaluated, so that the unmanned aerial vehicle can be widely applied. Especially, a deep learning algorithm is adopted for feature extraction and classification, and a dynamic threshold change detection algorithm is adopted, so that the accuracy and efficiency of ecological checking are greatly improved.

The invention belongs to the technical field of rail traffic safety monitoring, and particularly relates to a rail foreign matter dark light detection method and device based on change detection. The method comprises the following steps: acquiring an image frame set of a track scene and corresponding alarm area information, wherein the image frame set comprises a background frame and a plurality of analysis frames; inputting the background frame and the analysis frame into an improved DSIFN change detection algorithm for comparative analysis, and obtaining a change area on the analysis frame; and filtering the change area on the analysis frame based on the alarm area information, retaining the change area on the corresponding track surface in the analysis frame, acquiring coordinate information of the change area, and outputting an alarm. According to the method, the track foreign matter intrusion event is captured through change detection, and an edge information enhancement structure and a self-adaptive weight pooling operator are added in a basic change detection algorithm DSIFN for optimization, so that the detection capability under a dark light condition is improved.

This invention discloses a low-power data transmission method and system for IoT terminals based on edge collaboration, relating to the field of IoT wireless communication technology. The method includes: an event-driven acquisition step that uses a change detection algorithm to determine data changes, triggering transmission only when the change exceeds an adaptive threshold; otherwise, entering a deep sleep mode; a transmission scheduling step that uses deep reinforcement learning to optimize transmit power, modulation and coding, and transmission time slots; an edge collaborative processing step that performs semantic analysis and redundancy fusion on multi-terminal data before batch forwarding; a group collaborative transmission step that achieves channel state sharing and energy-balanced proxy forwarding within a group; and a feedback optimization step that feeds back the transmission effect to the terminal to update parameters. This invention, through an end-edge-cloud collaborative architecture and a closed-loop optimization mechanism, reduces terminal standby power consumption to the microampere level, improves energy efficiency, and extends network lifetime.

The invention discloses a tunnel blasting lumpiness in-situ real-time detection method based on video measurement. The method is realized through the following steps: installing cameras and high-intensity light sources on a cab, a movable arm and a bucket rod of a deslagging excavator, and matching the cameras and the high-intensity light sources; system calibration is completed by using an orientation sensor, a calibration plate and tunnel contour information; a muck pile change video is automatically collected in the deslagging process; screening typical frame segments through a change detection algorithm, and marking change types and spatio-temporal information; the disturbed rocks are restored based on the mark information, and a muck pile multi-layer three-dimensional morphological model is constructed from the outside to the inside in combination with the excavation exposure sequence; and finally extracting a lumpiness grading curve, rock spatial distribution and muck pile shape information. The system correspondingly comprises a camera, a light source, an analysis processing component, a calibration component and the like. The method solves the problems that in the prior art, the internal lumpiness of the muck pile is difficult to obtain, the detection aging is poor, and the environmental adaptability is insufficient, and automatic, instant and comprehensive detection of the tunnel blasting in-situ lumpiness is achieved.

The present invention belongs to the field of computer image processing, and specifically relates to an image processing method for flood control monitoring of the river and lake chief system based on artificial intelligence. Cameras are deployed at key positions of rivers and lakes, and drones are used to collect images. The images are preprocessed by improving the low-light enhancement and noise reduction algorithms to improve the image quality in low-light and bad weather conditions. The superpixel segmentation technology and the image change detection algorithm are used. First, the gradient information of the image is calculated to adaptively adjust the weight coefficients, and then the features are extracted and classified. The changes are judged by comparing the regions at different time points, and the flood risk is predicted by combining time series analysis. Finally, through the big data platform combined with real-time monitoring information, emergency plans are formulated for different risk levels, and decision-making supports such as resource allocation and public warning suggestions are provided. After the flood, evaluation and summary can also be carried out to improve the flood control strategy. This method can effectively solve many problems of existing flood control monitoring, improve the intelligent level and efficiency, and ensure the safety of people's lives and property and the stable development of the region.

This invention discloses a method and system for optimizing drought emergency response plans based on multi-source data fusion. The method includes collecting multi-source data and historical emergency data from a preset area, and preprocessing the multi-source data and historical emergency data. The multi-source data includes meteorological data, soil data, crop data, and water conservancy data. The multi-source data is input into a lightweight multimodal fusion network to obtain fused data. A dual-index temporal abrupt change detection algorithm and a weakly supervised convolutional autoencoder are used to capture sudden drought features in the fused data to obtain sudden drought features. A machine learning-based drought emergency response plan model is constructed based on the sudden drought features. Predicted emergency data is obtained based on the drought emergency response plan model. Decision bias fit is obtained based on the historical emergency data and the predicted emergency data. Multi-objective reinforcement learning is used to dynamically optimize decisions based on the decision bias fit to obtain the optimized result.

This invention discloses an intelligent method for extracting cultivated land outflow patches, including: data collection and preprocessing; rapid sample generation based on existing data and manual collection, and storage of sample metadata and file data in database and file formats, respectively; adaptive modification and combination of appropriate semantic segmentation and change detection algorithms, introduction of fully trained general segmentation and general change detection pre-trained models, and training of iterative optimization networks to obtain optimally accurate non-cultivated land identification and cultivated land change detection models; extraction of suspected cultivated land outflow patches based on the optimally accurate models; batch regularization of suspected cultivated land outflow patches to obtain preliminary suspected cultivated land outflow patches; and verification of the correctness of the boundaries and attributes of the preliminary patches to obtain the final suspected cultivated land outflow patches. This invention can monitor cultivated land outflow in mountainous areas and quickly, efficiently, and accurately extract and update change patches and type information.

The application discloses a foreground target barrage anti-shielding method based on artificial intelligence, and comprises the following steps: S10, sampling original video at a preset frequency to obtain video stream images; S20, processing the video stream by using a shot change detection algorithm to determine whether the current video frame has shot change; S30, processing the current image by using a face detector to obtain current image target face information; S40, processing the current image by using an instance segmenter to obtain current image target instance segmentation information; S50, processing the current image by using a semantic segmenter to obtain current image target human body semantic segmentation information, wherein the human body semantic segmentation information is a one-dimensional feature map with the same width and height as the original image, and the value on each pixel represents the probability that the pixel is a human body; S60, processing the current image by using a depth estimator to obtain current image depth estimation information; and S70, sending the above information to a post-processing module for processing.

The invention discloses a domain generalization remote sensing image change detection method based on cross-domain consistency learning, and the method comprises the following steps: carrying out the geometric enhancement of obtained synthetic dual-phase remote sensing image data, and dividing an obtained enhanced synthetic data set into a training set and a test set; constructing a domain generalization remote sensing image change detection network based on semantics; selecting a change detection algorithm as a feature extractor to obtain an output feature of an original image pair and an output feature of an enhanced image pair which are consistent with the output in spatial size, and respectively inputting the two groups of features into a prediction head of 3 * 3 convolution sharing weight to obtain two detection result images; constructing a space-time consistency comparative learning loss function and a pseudo change correction comparative learning loss function; and setting an overall loss function, iteratively training and optimizing network parameters, and inputting a detection image into the trained neural network after the loss is stable to obtain a final detection result graph. According to the method, the influence of similar category confusion caused by domain offset can be effectively reduced, so that the generalization ability of an existing change detection neural network model is effectively improved.

This invention addresses the technical problems of existing unsupervised change detection algorithms, which mainly rely on algebraic transformations and spectral unmixing of spectra, resulting in poor performance and limited applicability in hyperspectral data with mixed pixels. It provides a hyperspectral image change detection method based on an unsupervised spectral unmixing neural network. The method includes the following steps: 1. Segmenting and mixing two hyperspectral images from different time phases to obtain hyperspectral image patch data, which is used as the training dataset; 2. Constructing a spectral unmixing network to generate endmember matrices and abundance matrices with image features; 3. Reconstructing the hyperspectral image patch data based on the endmember matrices and abundance matrices, and selecting reconstruction errors to train the spectral unmixing network; 4. Generating a change grayscale image based on the trained spectral unmixing network and the abundance matrix output by the abundance matrix generation module; 5. Binarizing the change grayscale image using a threshold segmentation algorithm to obtain the final hyperspectral image change detection result.

The application discloses a kind of based on deep learning's optical remote sensing image building change detection method, belong to remote sensing image change detection field.Firstly, optical remote sensing image data is obtained from high two and is preprocessed and change information label Label is marked;Then the image is input into the change detection model of the application, and the detection result map is output after calculation.The application effectively optimizes and improves the problems that the existing change detection method has lower detection accuracy, the detection edge is not complete, and the change detection algorithm has poor universality on different data sets.

The invention discloses a method for evaluating the performance of a topographic change detection algorithm. The method comprises the following steps: S1, obtaining a to-be-measured algorithm; s2, constructing a test set for evaluating algorithm precision; s3, constructing a test set for evaluating the anti-error capability of the algorithm; s4, constructing a test set for evaluating the anti-scale effect capability of the algorithm; s5, respectively inputting the three test sets generated in the steps S2 to S4 into a to-be-evaluated algorithm for detection to obtain a test result under each test set; s6, processing the test result of each test set obtained in the S5, and calculating to obtain the geometric mean of the average error, the geometric mean of the standard error and the geometric mean of the Moran index under each test set; s7, evaluating the comprehensive performance of the algorithm; and comprehensively processing the geometric mean value of the average error, the geometric mean value of the standard error and the geometric mean value of the Moran index in the step S6 to obtain a final score about the algorithm performance. According to the application of the method, the optimal topographic change detection algorithm can be optimized, and an earth surface deformation detection algorithm is provided for the fields of disaster detection and the like.

The invention relates to the technical field of mining area monitoring, in particular to a mining area fixed-point data surveying, mapping, collecting and monitoring device which comprises a main control module used for coordinating and controlling operation of the device. The system comprises a multi-source data acquisition module, a data fusion processing module, a change detection algorithm module, a communication module, a positioning module, a power supply module, a storage module, an environment sensor module and a mobile terminal interface module. By integrating the satellite remote sensing unit, the unmanned aerial vehicle aerial photography unit and the ground measurement unit, a sky-ground integrated monitoring system is constructed, the limitation of a single data source can be overcome, mining area global coverage and high-frequency data acquisition are realized, and the monitoring efficiency and timeliness are remarkably improved.

The invention discloses a passive target positioning method combining Beidou navigation and remote sensing image change detection, which has the advantages that a remote sensing image change detection algorithm based on a global hypergraph is designed, and the algorithm takes an image block as a vertex and takes a global similar image block set as a hyperedge; constructing a global hypergraph model with image detail and global context representation capability; then, global hypergraph models are constructed on the remote sensing images at different moments respectively, measurement of change information is completed by matching high-order structure information and hyperedge affinity between hypergraphs, and then a binary change graph between the remote sensing images is obtained. In addition, the invention further provides a passive target positioning strategy, the center of a change connected domain in a binary change graph is solved, a position change reference point of an active target is generated, and a longitude and latitude grid under an image plane is constructed by means of a reference point geographic coordinate provided by a Beidou system. And the positions of other passive ground object targets in the image plane are accurately estimated.

The present invention proposes an instance-level change detection method and system for collaborative optimization of contour accuracy and positioning accuracy. Among them, the method includes: making change detection labels for preprocessed satellite data; training a change detection model using a joint loss composed of pixel-level loss and instance-level loss with the significance of positioning accuracy constraint; adapting the instance-level evaluation index AP10 to the evaluation of change detection results; evaluating the trained change detection model using the adapted evaluation index AP10; and inputting the initial change patches and dual-temporal images obtained by the trained change detection model into the SAM change detection algorithm to obtain change patches. The solution proposed by the present invention can meet the dual requirements of high positioning accuracy and high contour accuracy, improve the integrity of large patches, reduce the false detection and missed detection of small patches, enhance the boundary fit degree between the detected area and the real change area, and improve the contour accuracy and positioning accuracy of change detection results in complex scenes.

The invention discloses a subscription type remote sensing change monitoring method based on a large-scale constellation, belongs to the technical field of satellite technologies and remote sensing application services, solves the technical problems of a traditional change monitoring mode, and can meet the customized change monitoring requirement of a user in a large-area and high-frequency requirement scene. Generating a reference period remote sensing image of the subscription area; when the satellite accords with the shooting condition determined by the shooting decision model, scheduling the satellite to shoot the subscription area to obtain effective remote sensing image data; generating a change probability graph based on the reference period remote sensing image of the subscription area, the effective remote sensing image data after quality inspection and a deep learning change detection algorithm, and generating and outputting a to-be-pushed pattern spot based on the change probability graph; and repeating the above steps to produce the multi-expected to-be-pushed pattern spots, pushing the pattern spots meeting the standard based on the multi-expected to-be-pushed pattern spots and the variable pattern spot background database, and synchronously updating the variable pattern spot background database.

The present invention belongs to the technical field of high-resolution remote sensing image processing and remote sensing image data mining, and particularly relates to a remote sensing image change detection method and its detection system based on dynamic weighted cross-entropy loss. Step 1: Make a change detection data set based on high-resolution remote sensing images; Step 2: Construct a remote sensing image change detection model based on a weight-sharing Siamese neural network; Step 3: Design a dynamic weighted cross-entropy loss function and an optimizer; Step 4: Use the high-resolution remote sensing change detection data set to train the improved deep learning model; Step 5: Use data augmentation during testing to predict the test data set and perform post-processing operations to improve the prediction quality. The present invention is used to solve the problem of the deficiencies of remote sensing image change detection algorithms.

The present invention discloses a large-scale remote sensing fragmented farmland change detection algorithm based on coarse screening and fine inspection, comprising the following steps: Step S1: Acquire two large-scale farmland remote sensing images of the same location at different times and preprocess them separately; then crop the preprocessed images to obtain a pre-phase sub-image group and a post-phase sub-image group; Step S2: Select a histogram standardization method to unify the style of the pre-phase sub-image group and the post-phase sub-image group; Step S3: Coarsely screen the farmland change images to obtain a new image dataset, wherein the farmland change images are the pre-phase sub-image group and the post-phase sub-image group; Step S4: Finely detect and locate the farmland change areas in the new image dataset to obtain the farmland change status. This algorithm solves the problems of existing change detection methods used to detect farmland changes, such as the difficulty in obtaining label information, the small number of farmland change areas and the difficulty in detecting them, and the complexity of image acquisition and application.

The invention belongs to the technical field of rail traffic safety monitoring, and particularly relates to a rail foreign matter small target detection method and device based on change detection. The method comprises the following steps: acquiring an image frame set of a track scene and corresponding alarm area information, wherein the image frame set comprises a background frame and a plurality of analysis frames; inputting the background frame and the analysis frame into an improved SNUnet change detection algorithm for comparative analysis, and obtaining a change area on the analysis frame; and filtering the change area on the analysis frame based on the alarm area information, retaining the change area on the corresponding track surface in the analysis frame, acquiring coordinate information of the change area, and outputting an alarm. On the basis of an optimized SNUnet change detection algorithm, the problem that foreign matter invasion categories cannot be exhaustive can be solved, the foreign matter capturing capacity for small targets is improved, and the analysis performance can be further improved on the premise that the receptive field is expanded.

Systems and methods include mitigating interference in satellite communications, using predicted in-line event (ILE) detection between satellites of different systems operating on the same frequency, each prediction being generated via change detection algorithms based on machine learning models or ephemeris-based predictions so as to anticipate reductions in signal quality before they occur. Upon detection, proactive adaptive coding and modulation (P-ACM) adjusts signal characteristics in accordance with the predicted ILEs, the adjustments being sufficient to mitigate interference prior to degradation. Additionally, successive interference cancellation (SIC) and cooperative spectrum sharing techniques are employed to further enhance performance, wherein the combination of P-ACM, SIC, and cooperative spectrum sharing is sufficient to improve signal reliability, reduce latency, and optimize spectral efficiency in dense satellite networks, particularly in low Earth orbit (LEO) systems.

The application discloses a remote sensing image change detection algorithm based on multi-dimensional feature collaborative modeling. The steps are as follows: image block preprocessing and division of remote sensing image change detection public reference data set; a remote sensing image change detection model based on multi-dimensional feature collaborative modeling is constructed, for the output from the feature extraction stage, the main branch inputs it into a double-branch feature difference modeling module to model the difference of each layer of double-time feature pairs to form multi-scale difference features, and then inputs it into a boundary protection feature calibration module to strengthen the edge, and the enhanced features are input into a decoder to generate a detection result; the auxiliary branch inputs it into a style difference feature alignment module; after the model is trained, the test set is tested. On the four evaluation indexes of precision, recall, F1 score and intersection over union, compared with the existing change detection method FC-Siam-Conv, it is respectively increased by 3.72%, 6.93%, 5.18% and 8.38%; compared with the existing change detection method BIT, it is respectively increased by 1.63%, 1.11%, 1.38% and 2.32%.

The present application relates to a kind of electronic chart incremental updating file generation method, belong to marine surveying and mapping technical field, by constructing "EN object record table" to store original data, "ER object record table" records change information, in combination with topological influence diffusion analysis, realize the cascade change identification of point, curve, combination geometry and associated elements.Especially, for complex curve editing, a geometric change detection algorithm based on "node sequence double-pointer matching and tolerance threshold judgment" is designed, which can accurately identify the insertion, deletion and coordinate update operation of nodes, and mark the change type in the ER record table.This method effectively improves the automation, accuracy and traceability of electronic chart updating, and ensures the timeliness of navigation data.

The invention discloses a domain generalization remote sensing image change detection method based on domain invariant feature extraction, and the method comprises the following steps: carrying out the geometric enhancement of obtained synthetic dual-phase remote sensing image data, and dividing an obtained enhanced synthetic data set into a training set and a test set; constructing a domain generalization remote sensing image change detection network based on semantics; selecting a change detection algorithm as a feature extractor to obtain an output feature of an original image pair and an output feature of an enhanced image pair which are consistent with the output in spatial size, and respectively inputting the two groups of features into a prediction head of 3 * 3 convolution sharing weight to obtain two detection result images; constructing an auto-covariance matching loss function and a cross-covariance diagonal loss function; and setting an overall loss function, iteratively training and optimizing network parameters, and inputting a detection image into the trained neural network after the loss is stable to obtain a final detection result graph. According to the method, the domain offset between the training domain data set and the unseen domain data set can be effectively reduced, so that the generalization ability of an existing convolutional neural network model is effectively improved.