Fog cluster identification method, device, electronic device and storage medium
By preprocessing, global feature extraction and classification processing of scene images acquired by the camera, the problem of cluster fog recognition in the prior art is easily affected by light and angle, and higher recognition accuracy is achieved.
Patent Information
- Application Number
- CN202010699922.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2020-07-17
- Publication Date
- 2025-02-21
- Estimated Expiration
- 2040-07-17
AI Technical Summary
When using cameras to identify fog, the prior art is susceptible to factors such as light and camera angle, resulting in poor recognition effect.
By acquiring the scene image of the target scene, image preprocessing, global feature extraction and classification processing are performed to determine the fog and its level in the scene image. Specific steps include pixel sampling, normalization processing, global feature extraction and average pooling classification.
It improves the effectiveness of mass fog information representation, reduces information interference in the image that is not related to mass fog recognition, and thus improves the accuracy of mass fog recognition.
Smart Images
Figure CN111783732B_ABST
Abstract
Description
Technical Field
[0001] The present disclosure relates to the field of computer vision technology, and in particular to a method, device, electronic device and storage medium for identifying fog clusters. Background Art
[0002] "Cluster fog", also known as clumps of fog, is a type of fog with lower visibility that occurs in a local area of tens to hundreds of meters in the fog due to the influence of the local microclimate environment. Cluster fog is highly regional and difficult to predict. On highways, cluster fog can cause sudden changes in visibility, which is extremely harmful to highway traffic safety and can easily lead to major traffic accidents.
[0003] At present, dedicated hardware such as photoelectric sensors or cameras can be used to identify fog clusters. Compared with the high cost and difficult to promote large-scale solutions using dedicated hardware, using cameras to obtain image information and thus identify fog clusters is more feasible.
[0004] However, solutions for fog group recognition using cameras often use methods such as dark channel priors or manually selected features to obtain shallow image features from images captured by the camera. These methods are easily affected by factors such as light and camera angle, resulting in poor recognition results. Summary of the invention
[0005] The embodiments of the present disclosure are intended to provide a method and device for identifying fog clusters, an electronic device, and a storage medium.
[0006] The technical solution of the embodiment of the present disclosure is implemented as follows:
[0007] The present disclosure provides a method for identifying fog clusters, the method comprising:
[0008] Acquire a scene image of a target scene;
[0009] Extracting features from the scene image to obtain global feature information;
[0010] The global feature information is classified and processed to determine the fog clusters in the scene image and the fog cluster levels corresponding to the fog clusters.
[0011] In the above-mentioned fog group identification method, before extracting features from the scene image to obtain global feature information, the method further includes:
[0012] Performing image preprocessing on the scene image to obtain a processed image;
[0013] Accordingly, the step of extracting features from the scene image to obtain global feature information includes:
[0014] Feature extraction is performed on the processed image to obtain the global feature information.
[0015] In the above-mentioned fog group identification method, the image preprocessing of the scene image to obtain a processed image includes:
[0016] Performing pixel sampling on the scene image to obtain a sampled image of a target size;
[0017] The sampled image is normalized to obtain the processed image.
[0018] In the above-mentioned fog cluster identification method, the classification processing of the global feature information to determine the fog clusters in the scene image and the fog cluster levels corresponding to the fog clusters includes:
[0019] Performing average pooling on the global feature information to obtain global average feature information;
[0020] The global average characteristic information is classified to determine the fog clusters and the fog cluster levels.
[0021] In the above-mentioned fog group identification method, after determining the fog group in the scene image and the fog group level corresponding to the fog group, the method further includes:
[0022] According to the corresponding relationship between the preset fog group level and the visibility range, obtaining a first visibility range corresponding to the fog group level;
[0023] When the first visibility range is lower than a preset visibility range, obtaining first warning information corresponding to the fog group level;
[0024] The first warning information is output.
[0025] In the above-mentioned fog group identification method, the scene image includes multiple images at multiple moments, the fog group level includes multiple levels corresponding to the multiple images, and after determining the fog group in the scene image and the fog group level corresponding to the fog group, the method further includes:
[0026] When the multiple levels are positively correlated with the multiple moments, obtaining second warning information;
[0027] Acquire the early warning reminder device in the target scene;
[0028] The second warning information is transmitted to the warning reminder device.
[0029] In the above-mentioned fog group identification method, the scene image includes multiple images at multiple moments, the fog group level includes multiple levels corresponding to the multiple images, and after determining the fog group in the scene image and the fog group level corresponding to the fog group, the method further includes:
[0030] Performing statistics on the multiple levels to obtain statistical results;
[0031] The frequency of occurrence of fog clusters in the target scene is determined according to the statistical results.
[0032] The present disclosure provides a fog group identification device, the fog group identification device comprising:
[0033] An image acquisition module, used for acquiring a scene image of a target scene;
[0034] A feature extraction module, used to extract features from the scene image to obtain global feature information;
[0035] The classification processing module is used to classify the global feature information to determine the fog clusters in the scene image and the fog cluster levels corresponding to the fog clusters.
[0036] The above-mentioned fog group identification device also includes the image processing module.
[0037] The image processing module is used to perform image preprocessing on the scene image to obtain a processed image;
[0038] Correspondingly, the feature extraction module is specifically used to extract features from the processed image to obtain the global feature information.
[0039] In the above-mentioned fog group identification device, the image processing module is specifically used to perform pixel sampling on the scene image to obtain a sampled image of a target size; and perform normalization processing on the sampled image to obtain the processed image.
[0040] In the above-mentioned fog cluster identification device, the classification processing module is specifically used to perform average pooling on the global feature information to obtain global average feature information; and classify the global average feature information to determine the fog cluster and the fog cluster level.
[0041] The above-mentioned fog group identification device also includes an information output module.
[0042] The information output module is used to obtain a first visibility range corresponding to the fog group level according to a corresponding relationship between the preset fog group level and the visibility range; when the first visibility range is lower than the preset visibility range, obtain first warning information corresponding to the fog group level; and output the first warning information.
[0043] In the above-mentioned fog group identification device, an information output module is also included, the scene image includes a plurality of images at a plurality of moments, the fog group level includes a plurality of levels corresponding to the plurality of images,
[0044] The information output module is used to obtain second warning information when the multiple levels are positively correlated with the multiple moments; obtain the warning reminder device in the target scene; and transmit the second warning information to the warning reminder device.
[0045] In the above-mentioned fog group identification device, a statistical processing module is also included, the scene image includes a plurality of images at a plurality of moments, the fog group level includes a plurality of levels corresponding to the plurality of images,
[0046] The statistical processing module is used to perform statistics on the multiple levels to obtain statistical results; and determine the frequency of occurrence of fog clusters in the target scene based on the statistical results.
[0047] The embodiment of the present disclosure provides an electronic device, characterized in that the electronic device comprises: a processor, a memory and a communication bus; wherein:
[0048] The communication bus is used to realize the connection and communication between the processor and the memory;
[0049] The processor is used to execute the fog group identification program stored in the memory to implement the above-mentioned fog group identification method.
[0050] An embodiment of the present disclosure provides a computer-readable storage medium, which stores one or more programs. The one or more programs can be executed by one or more processors to implement the above-mentioned fog cluster identification method.
[0051] The embodiments of the present disclosure provide a method, device, electronic device and storage medium for identifying fog clusters. The method includes: acquiring a scene image of a target scene, extracting features from the scene image, and obtaining global feature information; classifying and processing the global feature information to determine the fog clusters in the scene image and the fog cluster levels corresponding to the fog clusters. The technical solution provided by the embodiments of the present disclosure extracts global deep-level features from the image, improves the effectiveness of representing fog cluster information, reduces the interference of information in the image that is irrelevant to fog cluster identification, and thus improves the accuracy of fog cluster identification. BRIEF DESCRIPTION OF THE DRAWINGS
[0052] Figure 1 A schematic diagram of a flow chart of a fog cluster identification method provided in an embodiment of the present disclosure;
[0053] Figure 2A schematic diagram of an exemplary image processing process provided by an embodiment of the present disclosure;
[0054] Figure 3 A schematic diagram of a fog group identification device provided in an embodiment of the present disclosure Figure 1 ;
[0055] Figure 4 A schematic diagram of a fog group identification device provided in an embodiment of the present disclosure Figure 2 ;
[0056] Figure 5 A schematic diagram of the structure of an electronic device provided in an embodiment of the present disclosure. DETAILED DESCRIPTION
[0057] The technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure.
[0058] The embodiment of the present disclosure provides a method for identifying a fog cluster, and the execution subject thereof may be a fog cluster identification device. For example, the method for identifying a fog cluster may be executed by a terminal device or a server or other electronic device, wherein the terminal device may be a user equipment (UE), a mobile device, a user terminal, a terminal, a cellular phone, a cordless phone, a personal digital assistant (PDA), a handheld device, a computing device, a vehicle-mounted device, a wearable device, etc. In some possible implementations, the method for identifying a fog cluster may be implemented by a processor calling a computer-readable instruction stored in a memory.
[0059] The embodiment of the present disclosure provides a method for identifying fog clusters. Figure 1 The following is a flow chart of a method for identifying fog clusters provided by an embodiment of the present disclosure. Figure 1 As shown in FIG. 1 , the cluster fog identification method mainly includes the following steps:
[0060] S101: Acquire a scene image of a target scene.
[0061] In the embodiments of the present disclosure, the fog group identification device can obtain a scene image of the target scene.
[0062] It should be noted that in the embodiments of the present disclosure, the scene image is the scene that needs to be identified by fog clusters, that is, the image corresponding to the target scene. Among them, when the fog cluster identification method is applied to the scene of highway monitoring, the scene image is actually the highway monitoring image, and the subsequent fog cluster identification is not limited by dedicated hardware, which is convenient for large-scale application. The specific target scene and the scene image of the target scene are not limited in the embodiments of the present disclosure.
[0063] It should be noted that in the embodiments of the present disclosure, the scene image may be collected by the fog group recognition device, or may be acquired by an independent camera, server, or cloud device and transmitted to the fog group recognition device. The specific source of the scene image is not limited in the embodiments of the present disclosure.
[0064] S102: Extract features from the scene image to obtain global feature information.
[0065] In the embodiment of the present disclosure, after obtaining the scene image, the fog group recognition device can extract features of the scene image, thereby obtaining global feature information in the scene image.
[0066] It should be noted that in the embodiments of the present disclosure, before the fog group identification device performs feature extraction on the scene image to obtain global feature information, it can also perform the following steps: perform image preprocessing on the scene image to obtain a processed image; accordingly, the fog group identification device performs feature extraction on the scene image to obtain global feature information, including: perform feature extraction on the processed image to obtain global feature information.
[0067] Specifically, in an embodiment of the present disclosure, the fog group identification device performs image preprocessing on the scene image to obtain a processed image, including: performing pixel sampling on the scene image to obtain a sampled image of a target size; and performing normalization processing on the sampled image to obtain a processed image.
[0068] It should be noted that in the embodiments of the present disclosure, the fog group identification device samples pixels of the scene image, which is actually to reshape the size of the scene image to a specific size, that is, the target size, so as to determine the obtained image as the sampled image. Among them, the fog group identification device can use a bilinear interpolation algorithm to sample the pixels of the scene image. Of course, other sampling algorithms can also be used to implement pixel sampling to obtain a sampled image. The specific target size and the pixel sampling method can be selected according to actual needs, and the embodiments of the present disclosure are not limited.
[0069] It should be noted that in the embodiment of the present disclosure, the fog group identification device performs normalization processing on the sampled image, which is essentially to convert the pixel value of each pixel in the sampled image into a numerical value between 0 and 1, thereby mapping the pixel values of different pixels into the same fixed range.
[0070] It can be understood that in the embodiments of the present disclosure, the fog group identification device performs pixel sampling and normalization processing on the scene image, and the obtained processed image can meet certain standards in terms of size and pixel representation, thereby facilitating subsequent feature extraction and improving the efficiency of feature extraction.
[0071] It should be noted that in the embodiments of the present disclosure, a preset deep neural network is stored in the fog group recognition device. The preset deep neural network can realize the extraction of global deep features in the image. The specific preset deep neural network can be obtained by training feature extraction using a large number of sample images, which is not limited in the embodiments of the present application.
[0072] It should be noted that in the embodiments of the present disclosure, the preset deep neural network may include: multiple groups of convolutional layers connected in sequence from shallow to deep according to the feature extraction depth. The fog group recognition device can use the multiple groups of convolutional layers to perform multiple iterative convolution processing on the processed image to obtain a feature image, and then determine the feature image as global feature information.
[0073] Figure 2 FIG. 1 is a schematic diagram of an exemplary image processing process provided by an embodiment of the present disclosure. Figure 2 As shown in the figure, the preset deep neural network includes 5 groups of convolutional layers, namely the first group of convolutional layers, the second group of convolutional layers, the third group of convolutional layers, the fourth group of convolutional layers and the fifth group of convolutional layers. Among them, the first group of convolutional layers uses a filter of 7×7 size, which can convolve the input 2-channel processed image to obtain a 64-channel image, and the image size is downsampled to 1 / 2; the second to fifth groups of convolutional layers, each group contains 1-3 layers of convolutional layers, the convolution kernel size is 3×3, and the number of convolution output channels is 64, 128, 256 and 512 respectively. The convolution result output by each group of convolutional layers, that is, the output image can be changed to 1 / 4, 1 / 8, 1 / 16 and 1 / 32 of the processed image. After the processed image is processed by 5 groups of convolutional layers, the image size gradually decreases, the channel gradually deepens, and the feature level becomes more advanced.
[0074] It should be noted that in the embodiments of the present disclosure, each group of convolutional layers can perform maximum pooling, that is, downsampling, on the image obtained by the convolution processing before outputting it to the next connected group of convolutional layers. Specifically, every four pixels in the image are grouped together, and only the pixel with the largest pixel value is retained, so that the length and width of the downsampled image are actually 1 / 2 of the input image.
[0075] It should be noted that in the embodiments of the present disclosure, the number of groups of convolutional layers included in the preset deep neural network, and the relevant parameters in each group of convolutional layers, such as the size of the convolution kernel, can be set according to actual needs, and the embodiments of the present disclosure are not limited thereto.
[0076] It can be understood that in the embodiments of the present disclosure, the global feature extraction of the processed image can be achieved by using a preset deep neural network. Compared with the traditional scheme of extracting local features of the image, it can more effectively characterize the fog information in the image, thereby improving the accuracy of fog recognition.
[0077] S103: Classify the global feature information to determine the fog clusters in the scene image and the fog cluster levels corresponding to the fog clusters.
[0078] In the embodiment of the present disclosure, after obtaining the global feature information of the scene image, the fog cluster identification device can classify the global feature information to determine the fog clusters in the scene image and the fog cluster levels corresponding to the fog clusters.
[0079] Specifically, in an embodiment of the present disclosure, a fog cluster identification device classifies and processes global feature information to determine fog clusters in a scene image and fog cluster levels corresponding to the fog clusters, including: performing average pooling on the global feature information to obtain global average feature information; classifying the global average feature information to determine fog clusters and fog cluster levels.
[0080] It should be noted that, in the embodiments of the present disclosure, Figure 2 As shown, the group fog recognition device may store a global pooling layer and a preset classifier, and the group fog recognition device may use the global pooling layer to realize the average pooling of global feature information. The global pooling layer may be set in a preset classifier, and in fact, may also be set in a preset deep neural network, and the function realized by it remains unchanged.
[0081] It should be noted that, in the embodiments of the present disclosure, Figure 2 As shown, the preset classifier may include a first fully connected layer, a second fully connected layer and a normalization layer, which are used to implement the classification of global average feature information. Among them, there may be two fully connected layers, of course, there may also be one or more than two. The specific number of fully connected layers can be set according to actual needs, and the embodiment of the present disclosure is not limited thereto.
[0082] It can be understood that in the embodiments of the present disclosure, since the global feature information is actually a feature image obtained by convolution processing, the fog group identification device uses a global pooling layer to perform global average pooling on the global feature information, that is, taking the average of each channel in the feature image, thereby avoiding the influence of interference factors in local areas of the image on fog group identification due to poor local light during imaging, thereby improving the accuracy of fog group identification.
[0083] It should be noted that in the embodiments of the present disclosure, the fog cluster identification device can classify the global average feature information to determine the fog clusters in the scene image and the fog cluster levels corresponding to the fog clusters. The global average feature information represents the average features of the scene image, and the fog cluster level actually represents the fog cluster situation of the scene image.
[0084] It should be noted that, in the embodiment of the present disclosure, the fog cluster identification device actually identifies the fog clusters based on the global average feature information, and classifies the fog clusters into fog cluster levels, thereby obtaining the proportions of different fog cluster levels, such as Figure 2 As shown in the figure, the global average feature information matches the third level of the fog cluster level the most, so the final determined fog cluster level is level 3. In addition, the output fog cluster level corresponds to visibility, which can more clearly and flexibly reflect the degree of fog cluster compared to the current fog cluster identification scheme, which outputs the results of whether there is fog cluster or the concentration of fog cluster.
[0085] Exemplarily, in an embodiment of the present disclosure, in the correspondence between the preset fog level and the visibility range, the visibility corresponding to fog level one is 0-50 meters, the visibility corresponding to fog level two is 50-100 meters, the visibility corresponding to fog level three is 100-200 meters, the visibility corresponding to fog level four is 200-500 meters, the visibility corresponding to fog level five is 500-1000 meters, and the visibility corresponding to fog-free is above 1000 meters.
[0086] In an embodiment of the present disclosure, after determining the fog clusters in the scene image and the fog cluster levels corresponding to the fog clusters, the fog cluster identification device can also perform the following steps: obtaining a first visibility range corresponding to the fog cluster level according to a preset correspondence between the fog cluster level and the visibility range; when the first visibility range is lower than the preset visibility range, obtaining first warning information corresponding to the fog cluster level; and outputting the first warning information.
[0087] It should be noted that in the embodiments of the present disclosure, a correspondence between preset fog levels and visibility ranges is stored, and different fog levels correspond to different visibility ranges. The specific correspondence between fog levels and visibility ranges can be pre-set according to actual conditions, and the embodiments of the present disclosure are not limited thereto.
[0088] It can be understood that in the embodiments of the present disclosure, the fog group identification device can search for the corresponding visibility range from the corresponding relationship between the preset fog group level and the visibility range according to the obtained fog group level, and determine the found visibility range as the first visibility range.
[0089] It should be noted that in the embodiment of the present disclosure, the fog group identification device stores a preset visibility range, for example, above 500 meters. When the fog group level obtained by the fog group identification device is between level 4 and level 1, the visibility represented by it is less than 500 meters. Therefore, it is necessary to determine the corresponding warning information. The specific warning information can be determined based on the corresponding relationship between the preset fog group level and the warning information, which is not limited in the embodiment of the present disclosure.
[0090] It should be noted that in the embodiments of the present disclosure, different fog levels correspond to different visibility levels. Therefore, when the visibility range corresponding to the fog level is lower than the preset visibility range, the corresponding warning information can be determined in time and output for prompting, thereby providing warning and scientific basis for the handling of fog.
[0091] In an embodiment of the present disclosure, the scene image includes multiple images at multiple moments, and the fog group level includes multiple levels corresponding to the multiple images. After the fog group identification device determines the fog group in the scene image and the fog group level corresponding to the fog group, the following steps can also be performed: when the multiple levels are positively correlated with the multiple moments, obtain the second warning information; obtain the warning reminder device in the target scene; and transmit the second warning information to the warning reminder device.
[0092] It can be understood that in the embodiments of the present disclosure, the fog group identification device can obtain a large number of images of the target scene within a period of time, for example, multiple images on a highway within a day, and the obtained images are all scene images, that is, the scene image can include multiple images at multiple moments. For each image, the fog group identification device can perform fog group identification separately, so as to obtain the corresponding fog group level, that is, multiple levels. If the multiple levels also show an improvement with the increase of time, that is, the multiple levels are positively correlated with the multiple moments, then the fog group characterizing the target scene is getting worse. The fog group identification device can obtain the early warning reminder device in the target scene, such as a vehicle, so as to transmit the second early warning information indicating that the fog group is getting worse to the early warning reminder device.
[0093] In an embodiment of the present disclosure, the scene image includes multiple images at multiple moments, and the fog group level includes multiple levels corresponding to the multiple images. After the fog group identification device determines the fog group in the scene image and the fog group level corresponding to the fog group, the following steps can also be performed: statistics are performed on the multiple levels to obtain statistical results; and based on the statistical results, the frequency of fog group occurrence in the target scene is determined.
[0094] It can be understood that in the embodiments of the present disclosure, the fog cluster identification device performs statistical analysis on the multiple levels corresponding to the multiple images obtained, and can determine the frequency of fog cluster occurrence in the target scene based on the statistical results. In addition, it can also analyze the frequent occurrence points and time periods of fog clusters, summarize the laws of fog cluster occurrence, and generate a fog cluster prevention map, so that prevention can be carried out in advance at the frequent occurrence points and time periods to reduce the losses caused by fog clusters.
[0095] The embodiment of the present disclosure provides a method for identifying a fog cluster, comprising: obtaining a scene image of a target scene; performing feature extraction on the scene image to obtain global feature information; and classifying and processing the global feature information to determine the fog clusters in the scene image and the fog cluster levels corresponding to the fog clusters. The method for identifying a fog cluster provided by the embodiment of the present disclosure uses a deep neural network to extract global features of an image, thereby improving the effectiveness of representing fog cluster information, reducing interference from information in the image that is irrelevant to fog cluster identification, and thus improving the accuracy of fog cluster identification.
[0096] The present disclosure provides a fog cluster identification device. Figure 3 A schematic diagram of a fog group identification device provided in an embodiment of the present disclosure Figure 1 .like Figure 3 As shown, the group fog identification device includes:
[0097] An image acquisition module 201 is used to acquire a scene image of a target scene;
[0098] A feature extraction module 202 is used to extract features from the scene image to obtain global feature information;
[0099] The classification processing module 203 is used to classify the global feature information to determine the fog clusters in the scene image and the fog cluster levels corresponding to the fog clusters.
[0100] Figure 4 A schematic diagram of a fog group identification device provided in an embodiment of the present application Figure 2 .like Figure 4 As shown, in one embodiment of the present disclosure, the fog group identification device further includes: an image processing module 204,
[0101] The image processing module 204 is used to perform image preprocessing on the scene image to obtain a processed image;
[0102] Correspondingly, the feature extraction module 202 is specifically used to extract features from the processed image to obtain the global feature information.
[0103] In an embodiment of the present disclosure, the image processing module 204 is specifically configured to perform pixel sampling on the scene image to obtain a sampled image of a target size; and perform normalization processing on the sampled image to obtain the processed image.
[0104] In one embodiment of the present disclosure, the classification processing module 203 is specifically used to perform average pooling on the global feature information to obtain global average feature information; classify the global average feature information to determine the fog clusters and the fog cluster levels.
[0105] In one embodiment of the present disclosure, Figure 4 As shown, the fog group identification device also includes: an information output module 205, and the information output module 205 is used to obtain a first visibility range corresponding to the fog group level according to a preset correspondence between the fog group level and the visibility range; when the first visibility range is lower than the preset visibility range, obtain the first warning information corresponding to the fog group level; and output the first warning information.
[0106] In one embodiment of the present disclosure, the scene image includes a plurality of images at a plurality of moments, and the fog level includes a plurality of levels corresponding to the plurality of images.
[0107] The information output module 205 is used to obtain second warning information when the multiple levels are positively correlated with the multiple moments; obtain the warning reminder device in the target scene; and transmit the second warning information to the warning reminder device.
[0108] In one embodiment of the present disclosure, Figure 4 As shown, the group fog identification device further includes: a statistical processing module 206, the scene image includes a plurality of images at a plurality of moments, the group fog level includes a plurality of levels corresponding to the plurality of images,
[0109] The statistical processing module 206 is used to perform statistics on the multiple levels to obtain statistical results; and determine the frequency of occurrence of fog clusters in the target scene according to the statistical results.
[0110] The embodiment of the present disclosure provides a fog cluster recognition device, which obtains a scene image of a target scene; extracts the scene image to obtain global feature information; and classifies the global feature information to determine the fog clusters in the scene image and the corresponding fog cluster levels. The fog cluster recognition device provided by the embodiment of the present disclosure uses a deep neural network to extract global features of an image, thereby improving the effectiveness of fog cluster information representation, reducing the interference of information in the image that is irrelevant to fog cluster recognition, and thus improving the accuracy of fog cluster recognition.
[0111] An embodiment of the present disclosure provides an electronic device. Figure 5 The structure diagram of an electronic device provided by the embodiment of the present disclosure is shown in FIG. Figure 5 As shown, the electronic device includes: a processor 301, a memory 302 and a communication bus 303; wherein,
[0112] The communication bus 303 is used to realize the connection and communication between the processor 301 and the memory 302;
[0113] The processor 301 is used to execute the fog group recognition program stored in the memory 302 to implement the above-mentioned fog group recognition method.
[0114] The embodiment of the present disclosure provides a computer-readable storage medium, which stores one or more programs, and the one or more programs can be executed by one or more processors to implement the above-mentioned group fog identification method. The computer-readable storage medium can be a volatile memory (volatile memory), such as a random access memory (Random-Access Memory, RAM); or a non-volatile memory (non-volatile memory), such as a read-only memory (Read-Only Memory, ROM), a flash memory (flash memory), a hard disk (Hard Disk Drive, HDD) or a solid-state drive (Solid-State Drive, SSD); or it can be a respective device including one or any combination of the above-mentioned memories, such as a mobile phone, a computer, a tablet device, a personal digital assistant, etc.
[0115] Those skilled in the art will appreciate that the embodiments of the present disclosure may be provided as methods, systems, or computer program products. Therefore, the present disclosure may take the form of hardware embodiments, software embodiments, or embodiments combining software and hardware. Furthermore, the present disclosure 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 and optical storage, etc.) containing computer-usable program code.
[0116] The present disclosure is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present disclosure. 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 signal processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable signal 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.
[0117] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable signal processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture including 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.
[0118] These computer program instructions may also be loaded onto a computer or other programmable signal processing device so that a series of operational steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable device to implement 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.
[0119] The above description is only a preferred embodiment of the present disclosure and is not intended to limit the protection scope of the present disclosure.
Claims
1. A method for identifying fog clusters, characterized in that: The method comprises: Acquire a scene image of a target scene; Extracting features from the scene image to obtain global feature information; Classify the global feature information to determine the fog clusters in the scene image and the fog cluster levels corresponding to the fog clusters; The step of extracting features from the scene image to obtain global feature information includes: Performing multiple iterative convolution processing on the scene image using a preset deep neural network to obtain a feature image, and determining the feature image as the global feature information; the preset deep neural network is used to extract global deep-level features from the scene image; The classifying and processing the global feature information to determine the fog clusters in the scene image and the fog cluster levels corresponding to the fog clusters includes: Performing average pooling on the global feature information to obtain global average feature information; The global average feature information is classified using a preset classifier to determine the fog cluster and the fog cluster level.
2. The method according to claim 1, characterized in that Before extracting features from the scene image to obtain global feature information, the method further includes: Performing image preprocessing on the scene image to obtain a processed image; Accordingly, the step of extracting features from the scene image to obtain global feature information includes: Feature extraction is performed on the processed image to obtain the global feature information.
3. The method for identifying fog clusters according to claim 2, characterized in that: The performing image preprocessing on the scene image to obtain a processed image includes: Performing pixel sampling on the scene image to obtain a sampled image of a target size; The sampled image is normalized to obtain the processed image.
4. The method for identifying fog clusters according to any one of claims 1 to 3, characterized in that: After determining the fog clusters in the scene image and the fog cluster levels corresponding to the fog clusters, the method further includes: According to the corresponding relationship between the preset fog group level and the visibility range, obtaining a first visibility range corresponding to the fog group level; When the first visibility range is lower than a preset visibility range, obtaining first warning information corresponding to the fog group level; The first warning information is output.
5. The method for identifying fog clusters according to any one of claims 1 to 3, characterized in that: The scene image includes a plurality of images at a plurality of moments, the fog group level includes a plurality of levels corresponding to the plurality of images, and after determining the fog group in the scene image and the fog group level corresponding to the fog group, the method further includes: When the multiple levels are positively correlated with the multiple moments, obtaining second warning information; Acquire the early warning reminder device in the target scene; The second warning information is transmitted to the warning reminder device.
6. The method for identifying fog clusters according to any one of claims 1 to 3, characterized in that: The scene image includes a plurality of images at a plurality of moments, the fog group level includes a plurality of levels corresponding to the plurality of images, and after determining the fog group in the scene image and the fog group level corresponding to the fog group, the method further includes: Performing statistics on the multiple levels to obtain statistical results; The frequency of occurrence of fog clusters in the target scene is determined according to the statistical results.
7. A fog group identification device, characterized in that: The group fog identification device comprises: An image acquisition module, used for acquiring a scene image of a target scene; A feature extraction module, used to perform multiple iterative convolution processing on the scene image using a preset deep neural network to obtain a feature image, and determine the feature image as global feature information; the preset deep neural network is used to realize the extraction of global deep-level features in the scene image; The classification processing module is used to perform average pooling on the global feature information to obtain global average feature information; use a preset classifier to classify the global average feature information to determine the fog clusters in the scene image and the fog cluster levels corresponding to the fog clusters.
8. An electronic device, characterized in that: The electronic device comprises: a processor, a memory and a communication bus; wherein, The communication bus is used to realize the connection and communication between the processor and the memory; The processor is used to execute the fog group identification program stored in the memory to implement the fog group identification method described in any one of claims 1-6.
9. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores one or more programs, and the one or more programs can be executed by one or more processors to implement the fog group identification method according to any one of claims 1 to 6.
Citation Information
Patent Citations
Foggy weather visibility identification method and system under complex background
CN110378865A