A rescue method, device, system and equipment based on identifying a light distress signal

Abstract

The application discloses a kind of based on identification light distress signal rescue method, device, system and equipment, this method includes steps: rescue system obtains the light signal video image of suspected person in distress that mobile terminal is shot, then suspected person in distress light signal video image is input light distress signal dynamic model, judge whether it is light help signal video image after: if suspected person in distress light signal video image is light help signal video image, then according to light help signal video image and the position of mobile terminal, determine the emission position of light help signal;Light help signal video image and the emission position of light help signal are alarmed and issued, to make professional person come to rescue person in distress, to realize the rescue based on video dynamic analysis identification distress signal.This method can be based on video dynamic analysis and identify distress signal, to rescue person in distress in residential building / building etc. building construction, and cannot communicate with outside or cannot send distress sound but can only use light to send distress signal.

Description

Technical Field

[0001] This invention relates to the field of network technology, and in particular to a rescue method, apparatus, system and equipment based on the identification of optical distress signals. Background Technology

[0002] The rapid development of video processing technology has driven the emergence and development of video recognition technology, which has gradually become an important component of the field of artificial intelligence. Video analysis focuses on studying the content of videos; it tends to analyze, interpret, and identify video content, and mark targets and trajectories for anomalies in video footage. Currently, video recognition has been widely applied in various fields, such as intrusion detection, queue management, and abnormal behavior detection.

[0003] However, in the current technology, there is no effective solution for people trapped in residential buildings or high-rise buildings who cannot communicate with the outside world or make a distress call by using lights to send out a distress signal. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to address the above-mentioned shortcomings of the prior art by proposing a rescue method, device, system and equipment based on the identification of light distress signals. The method can identify distress signals based on video dynamic analysis and then determine the transmission location of the light distress signal in order to rescue people trapped in residential buildings or high-rise buildings who are unable to communicate with the outside world or make distress sounds and can only use lights to make distress signals.

[0005] In a first aspect, the present invention provides a rescue method based on identifying optical distress signals, which is applied to a rescue system and includes the following steps:

[0006] Step S1: Acquire video images of light signals suspected to be from trapped individuals captured by the mobile terminal;

[0007] Step S2: After inputting the suspected trapped person's optical signal video image into the optical distress signal dynamic model, determine whether the suspected trapped person's optical signal video image is indeed an optical distress signal video image:

[0008] If the video image of the suspected trapped person is not a video image of a distress signal, repeat steps S1 to S2; if the video image of the suspected trapped person is a video image of a distress signal, determine the transmission location of the distress signal based on the video image of the distress signal and the location of the mobile terminal, and proceed to step S3.

[0009] Step S3: Issue an alarm by displaying the video image of the distress signal and the location of its transmission, thereby enabling rescue based on dynamic video analysis and identification of distress signals.

[0010] Further, in step S2, the suspected trapped person's optical signal video image is input into the optical distress signal dynamic model to determine whether the suspected trapped person's optical signal video image is indeed an optical distress signal video image. This specifically includes the following steps:

[0011] Step A1: Extract model features from the optical signal video images of the suspected trapped person to obtain model feature extraction data;

[0012] Step A2: Preprocess the model feature extraction data;

[0013] Step A3: Input the preprocessed model feature extraction data into the preset optical distress signal dynamic model to extract the feature vector of the optical signal video image of the suspected trapped person;

[0014] Step A4: Based on the feature vector of the trapped person's light signal video image, classify and predict the trapped person's light signal video image to determine whether the suspected trapped person's light signal video image is a light distress signal video image.

[0015] Furthermore, in step S2, the transmission location of the optical distress signal is determined based on the video image of the optical distress signal and the location of the mobile terminal, specifically including the following steps:

[0016] Step B1: Extract location image features from the video image of the distress signal;

[0017] Step B2: Match the extracted location image features with a publicly available or system cloud-based location database to obtain the possible transmission location of the optical distress signal;

[0018] Step B3: Perform a secondary matching of the possible transmission locations of the optical distress signal based on the location of the mobile terminal. After the secondary matching is successful, the transmission location of the optical distress signal is obtained.

[0019] in:

[0020] The location of the mobile terminal is determined based on the mobile terminal's positioning.

[0021] A successful secondary match is defined as the distance between the mobile terminal and the possible transmission location of the optical distress signal being within a preset range.

[0022] Furthermore, before step S1, there is also a step S0, which includes: constructing a dynamic model of the optical distress signal, which includes the following steps:

[0023] Step S01: Collect images of various possible historical distress signals;

[0024] Step S02: Extract key image features corresponding to historical distress signal images;

[0025] The key image features include image brightness numerical features, image color numerical features, image texture numerical features, and image shape numerical features;

[0026] Step S03: Input the historical distress signal images and their corresponding key image features into the initial machine learning algorithm to obtain the initial dynamic model of the optical distress signal;

[0027] Step S04: Train the initial optical distress signal dynamic model using training data, and fine-tune the initial optical distress signal dynamic model using test data to obtain the optical distress signal dynamic model.

[0028] Secondly, the present invention provides a rescue device based on identifying optical distress signals, which is applied to a rescue system and includes:

[0029] The acquisition unit is used to acquire video images of light signals from suspected trapped individuals captured by a mobile terminal.

[0030] The judgment unit, connected to the acquisition unit, is used to determine whether the suspected trapped person's optical signal video image is an optical distress signal video image after inputting it into the optical distress signal dynamic model.

[0031] A location determination unit, connected to the judgment unit, is used to determine the transmission location of the light distress signal based on the video image of the light distress signal and the location of the mobile terminal;

[0032] An alarm unit, connected to the location determination unit, is used to issue an alarm by displaying the video image of the light distress signal and the transmission location of the light distress signal, thereby enabling rescue based on dynamic video analysis and identification of distress signals.

[0033] Furthermore, the determination unit includes:

[0034] The first extraction module is used to extract model features from the optical signal video images of suspected trapped persons to obtain model feature extraction data.

[0035] A preprocessing module, connected to the extraction module, is used to preprocess the model feature extraction data;

[0036] The second extraction module is connected to the preprocessing module and is used to input the preprocessed model feature extraction data into the preset light distress signal dynamic model to extract the feature vector of the suspected trapped person's light signal video image.

[0037] The prediction module, connected to the second extraction module, is used to classify and predict the light signal video image of the trapped person based on the feature vector of the light signal video image of the trapped person, so as to determine whether the suspected light signal video image of the trapped person is a light distress signal video image.

[0038] Furthermore, the position determination unit includes:

[0039] The third extraction module is used to extract location image features from the video image of the light distress signal;

[0040] The first matching module, connected to the third extraction module, is used to match the extracted location image features with a publicly available or system cloud-based location database to obtain the possible transmission location of the light distress signal.

[0041] The second matching module, connected to the first matching module, is used to perform secondary matching on the possible transmission locations of the optical distress signal based on the location of the mobile terminal. After successful secondary matching, the transmission location of the optical distress signal is obtained.

[0042] in:

[0043] The location of the mobile terminal is determined based on the mobile terminal's positioning.

[0044] A successful secondary match is defined as the distance between the mobile terminal and the possible transmission location of the optical distress signal being within a preset range.

[0045] Furthermore, the device also includes a construction unit for constructing a dynamic model of an optical distress signal, the construction unit comprising:

[0046] The collection module is used to collect images of various possible historical distress signals;

[0047] The fourth extraction module, connected to the collection module, is used to extract key image features corresponding to historical distress signal images;

[0048] The key image features include image brightness numerical features, image color numerical features, image texture numerical features, and image shape numerical features;

[0049] The input module, connected to the fourth extraction module, is used to input historical distress signal images and their corresponding key image features into the initial machine learning algorithm to obtain an initial dynamic model of the light distress signal.

[0050] The optimization module, connected to the input module, is used to train the initial optical distress signal dynamic model using training data and optimize the initial optical distress signal dynamic model using test data to obtain the optical distress signal dynamic model.

[0051] Thirdly, the present invention provides a rescue system comprising a mobile terminal and a server, wherein the mobile terminal and the server are connected via a network.

[0052] The mobile terminal is used to capture video images of light signals from suspected trapped individuals and transmit these images to the server.

[0053] The server is used to input the suspected trapped person optical signal video image into the optical distress signal dynamic model after receiving the suspected trapped person optical signal video image transmitted by the mobile terminal, and determine whether the suspected trapped person optical signal video image is an optical distress signal video image. If it is determined to be an optical distress signal video image, the server determines the transmission position of the optical distress signal based on the optical distress signal video image and the position of the mobile terminal.

[0054] The server is also used to issue alarms by displaying video images of the light distress signal and the location of the light distress signal, thereby enabling rescue based on dynamic video analysis and identification of distress signals.

[0055] Fourthly, the present invention provides an electronic device comprising a memory and a processor, wherein the memory stores a computer program, and when the processor runs the computer program stored in the memory, the processor executes the rescue method based on identifying light distress signals as described in the first aspect.

[0056] The beneficial effects of this invention are:

[0057] (1) The present invention can identify distress signals based on video dynamic analysis, and then determine the transmission location of the light distress signal, so as to rescue people trapped in residential buildings or high-rise buildings who are unable to communicate with the outside world or make distress sounds but can only use lights to make distress signals.

[0058] (2) The present invention can assist rescue operations through video images of light signals. By using server-side analysis and judgment, it can quickly identify light distress signals and determine their transmission location, thereby providing guidance and support for rescue. Attached Figure Description

[0059] Figure 1 This is a schematic diagram of a rescue method based on identifying optical distress signals in an embodiment of the present invention;

[0060] Figure 2 This is a schematic diagram of the rescue process based on the identification of light distress signals in an embodiment of the present invention;

[0061] Figure 3 This is a schematic diagram of a rescue device based on recognizing optical distress signals in an embodiment of the present invention;

[0062] In the attached figures, the reference numerals are: 10, acquisition unit; 20, judgment unit; 30, location determination unit; and 40, alarm unit. Detailed Implementation

[0063] To enable those skilled in the art to better understand the technical solution of the present invention, the embodiments of the present invention will be further described in detail below with reference to the accompanying drawings.

[0064] It is understood that the specific embodiments and accompanying drawings described herein are merely for explaining the invention and are not intended to limit the invention.

[0065] It is understood that, without conflict, the various embodiments and features in the embodiments of the present invention can be combined with each other.

[0066] It is understood that, for ease of description, only the parts related to the present invention are shown in the accompanying drawings, while the parts unrelated to the present invention are not shown in the drawings.

[0067] It is understood that each unit or module involved in the embodiments of the present invention may correspond to only one entity structure, or may be composed of multiple entity structures, or multiple units or modules may be integrated into one entity structure.

[0068] It is understood that, without conflict, the functions and steps marked in the flowcharts and block diagrams of this invention may occur in a different order than that marked in the accompanying drawings.

[0069] It is understood that the flowcharts and block diagrams of this invention illustrate the possible architecture, functions, and operations of systems, apparatuses, devices, and methods according to various embodiments of this invention. Each block in the flowchart or block diagram may represent a unit, module, program segment, or code, containing executable instructions for implementing the specified function. Furthermore, each block or combination of blocks in the block diagram and flowchart can be implemented using a hardware-based system to achieve the specified function, or using a combination of hardware and computer instructions.

[0070] It is understood that the units and modules involved in the embodiments of the present invention can be implemented by software or by hardware. For example, the units and modules can be located in a processor.

[0071] Example 1:

[0072] like Figure 1 As shown, this invention provides a rescue method based on identifying optical distress signals. This method is applied to a rescue system and includes the following steps:

[0073] Step S1: Acquire video images of light signals from suspected trapped individuals captured by the mobile terminal;

[0074] Step S2: After inputting the suspected trapped person's optical signal video image into the optical distress signal dynamic model, determine whether the suspected trapped person's optical signal video image is indeed an optical distress signal video image:

[0075] If the video image of the suspected trapped person is not a video image of a distress signal, repeat steps S1 to S2; if the video image of the suspected trapped person is a video image of a distress signal, determine the transmission location of the distress signal based on the video image of the distress signal and the location of the mobile terminal, and proceed to step S3.

[0076] As a specific implementation method, in step S2, the suspected trapped person's optical signal video image is input into the optical distress signal dynamic model to determine whether the suspected trapped person's optical signal video image is an optical distress signal video image. This specifically includes the following steps:

[0077] Step A1: Extract model features from the optical signal video images of the suspected trapped person to obtain model feature extraction data;

[0078] Step A2: Preprocess the model feature extraction data;

[0079] Step A3: Input the preprocessed model feature extraction data into the preset optical distress signal dynamic model to extract the feature vector of the optical signal video image of the suspected trapped person;

[0080] Step A4: Based on the feature vector of the trapped person's light signal video image, classify and predict the trapped person's light signal video image to determine whether the suspected trapped person's light signal video image is a light distress signal video image.

[0081] Specifically, when processing optical signal video images, a recognition model can be used. For example, SOS has a fixed international standard frequency of illumination, or the rules can be set, such as three short, three long, and three short.

[0082] In one specific implementation, step S2 involves determining the transmission location of the optical distress signal based on the video image of the optical distress signal and the location of the mobile terminal. This includes the following steps:

[0083] Step B1: Extract location image features from the video image of the distress signal;

[0084] Location image feature extraction from video images of distress signals can help determine the transmission location of the distress signal. As a specific implementation method, location image feature extraction methods include: (1) Visual feature extraction: Visual features such as color, texture, and edges are extracted through image processing and computer vision algorithms. These features can be used to distinguish the image features of different locations. (2) Landmark recognition: The transmission location of the distress signal can be determined by identifying landmarks or iconic buildings in the image. This can be achieved by matching the landmarks in the image with a known landmark database. (3) Geographical location information: The transmission location of the distress signal can be determined by using the geographic location information of the image, such as latitude, longitude, and altitude. This can be obtained through GPS positioning information of mobile terminals or other positioning technologies. (4) Environmental background analysis: Analyzing the environmental background in the image, including terrain, vegetation, and water, can provide clues about the transmission location of the distress signal. For example, specific environmental backgrounds such as mountains and coastlines can help determine the transmission location of the distress signal.

[0085] Step B2: Match the extracted location image features with a publicly available or system cloud-based location database to obtain the possible transmission location of the optical distress signal;

[0086] Step B3: Perform a secondary matching of the possible transmission locations of the optical distress signal based on the location of the mobile terminal. After the secondary matching is successful, the transmission location of the optical distress signal is obtained.

[0087] in:

[0088] The location of the mobile terminal is determined based on the mobile terminal's positioning.

[0089] A successful secondary match indicates that the distance between the mobile terminal and the possible transmission location of the optical distress signal is within a preset range.

[0090] Location, position information, and shooting parameters can be obtained from information on network-connected shooting devices. Distance feature extraction can be performed by calculating based on the photographer's location and shooting parameters.

[0091] Step S3: Issue an alarm by displaying the video image of the distress signal and the location of its transmission, so that professionals can rescue the trapped person, thereby achieving rescue based on dynamic video analysis and identification of distress signals.

[0092] As one specific implementation, before step S1, there is also step S0, which is used to construct a dynamic model of the optical distress signal, and includes the following steps:

[0093] Step S01: Collect images of various possible historical distress signals;

[0094] Step S02: Extract key image features corresponding to historical distress signal images;

[0095] Step S03: Input the historical distress signal images and their corresponding key image features into the initial machine learning algorithm to obtain the initial dynamic model of the optical distress signal;

[0096] Key image features include image brightness numerical features, image color numerical features, image texture numerical features, and image shape numerical features.

[0097] The key image feature extraction in this embodiment uses one of the following methods:

[0098] (1) SIFT features: Scale-Invariant Feature Transform (SIFT) is a commonly used feature extraction algorithm. It represents image features by detecting key points in an image and calculating descriptors of these key points. SIFT features are robust to changes in image rotation, scaling, and illumination.

[0099] (2) SURF Features: Speeded Up Robust Features (SURF) is a feature extraction algorithm based on SIFT. It represents image features by detecting interest points in an image and computing descriptors for these interest points. SURF features are optimized for computational efficiency and are suitable for large-scale image matching.

[0100] (3) HOG Features: Histogram of Oriented Gradients (HOG) is a feature description method used for object detection and image recognition. It represents image features by calculating the gradient histograms in different directions of the image. HOG features have wide applications in fields such as human detection and pedestrian recognition.

[0101] (4) CNN Features: Convolutional Neural Network (CNN) is a deep learning model that has achieved significant breakthroughs in the field of image processing. By using a pre-trained CNN model, high-level semantic features of images can be extracted, such as object shape and texture.

[0102] Step S04: Train the initial optical distress signal dynamic model using training data, and fine-tune the initial optical distress signal dynamic model using test data to obtain the optical distress signal dynamic model.

[0103] To make this embodiment clearer, as Figure 2As shown, when a captive sends a distress signal using light, this embodiment can proceed as follows: When an external user captures video using their terminal, the identification and rescue system identifies the video content and compares it with a dynamic light distress signal model stored in the system's cloud. If the match is successful, it is determined to be a distress signal. Upon triggering the distress signal identification, a location image is captured and its features are extracted. The system then matches possible locations using a publicly available or cloud-based location database and utilizes the mobile terminal's location information, based on the video capture parameters and image distance features, to further confirm the location. By prompting the user to make a determination, the system helps contact professionals to assist the person in need.

[0104] The specific process is as follows:

[0105] (1) Terminal video shooting: External terminal users use terminal devices to shoot videos and use the videos as input data.

[0106] (2) Distress signal recognition: By recognizing the video content, a dynamic light distress signal model is used for comparison to determine whether a match is successful. If a match is successful, it is determined to be a distress signal.

[0107] (3) Capture location image and extract features: After triggering distress signal recognition, capture location image from video and extract features. These features may include information such as image brightness, color, and texture.

[0108] (4) Location matching: Using a public or cloud-based location database, the captured location image is matched with locations in the database. Possible locations are found by comparing features.

[0109] One method involves using a publicly available or cloud-based location database to match captured location images with locations in the database. The specific implementation method is as follows:

[0110] (4.1) Building a location database: First, a location database needs to be built, which contains image data and corresponding location information for each location. This image data can be collected and organized by professionals, or it can be a publicly available dataset of location images.

[0111] (4.2) Feature Extraction and Matching: For the captured location images, computer vision methods can be used for feature extraction, such as extracting information about the image's color, texture, and shape. Then, the extracted features are matched with images in the location database, using feature matching algorithms such as SIFT, SURF, and ORB.

[0112] (4.3) Matching results: Based on the results of the matching algorithm, the location image most similar to the captured location image can be obtained. At the same time, the location information corresponding to the location image can also be obtained.

[0113] (4.4) Location Confirmation: By matching with locations in the location database, the location corresponding to the captured location image can be initially determined. Combined with other methods (such as distance calculation, angle calculation, etc.), the location can be further confirmed.

[0114] In this embodiment, a diverse range of location images were collected to cover different scenes and angles when constructing the location database. Simultaneously, appropriate algorithms and parameters were selected during feature matching to improve matching accuracy.

[0115] (5) Location Confirmation: Using the location information of the mobile terminal, combined with the parameters of the video recording and the distance features of the image, the location is confirmed a second time. This can be done by calculating distance, angle, etc.

[0116] Among these methods, calculating distances and angles can help confirm the location. The following methods can be used:

[0117] (5.1) Distance Calculation: Using the location information of the mobile terminal, the distance between the terminal device and the distress signal can be calculated. This can be achieved through methods such as triangulation. Based on the distance calculation results, the relative positional relationship between the terminal device and the distress signal can be determined.

[0118] (5.2) Angle Calculation: Using the mobile terminal's sensors (such as gyroscopes, accelerometers, etc.), the relative angle between the terminal device and the distress signal can be calculated. By measuring the device's direction or rotation angle, the terminal device's orientation relative to the distress signal can be determined.

[0119] (5.3) Visual Feature Matching: Visual features, such as landmarks and buildings, can be extracted from the captured location images. By matching these features with images in a location database, similar locations can be found. This can also be used as a method for location confirmation.

[0120] (6) Prompt the end user to make a judgment: Based on the location confirmation result, prompt the end user to make a judgment. Provide a corresponding interface or instructions to help the end user contact professionals for assistance.

[0121] Specifically, based on the location confirmation result, the result can be displayed to the end user, and a corresponding interface or instructions can be provided to help the end user contact professionals for assistance. One specific implementation method of this embodiment is as follows:

[0122] (6.1) Prompt Interface: Displays the location confirmation result on the terminal device. It can show the location relationship between the terminal device and the distress signal in map form, or display location information matched with the location database. The interface can be marked with location information, distance, direction, etc., to help the terminal user make a judgment.

[0123] (6.2) Contact and Assistance Buttons: Provide corresponding buttons or options on the prompt interface to enable end users to quickly contact and assist professionals. By clicking the button or selecting the option, functions such as telephone calls, SMS sending, and sending distress signals can be triggered, so that end users can communicate and assist professionals in real time.

[0124] (6.3) Navigation Instructions: If the end user needs to go to a location near the confirmed location, navigation instructions can be provided to help the end user quickly reach the destination. Navigation instructions may include map navigation, text guidance, voice prompts, etc., to provide clear navigation information.

[0125] (6.4) Information Sharing: Share the location confirmation results and related information with professionals so that they can better understand the end-user's situation and take appropriate rescue measures. Information can be transmitted to professionals through network connections or messaging.

[0126] The specific interface and instruction design can be adjusted and optimized according to actual needs and user experience.

[0127] Example 2:

[0128] like Figure 3 As shown, this embodiment provides a rescue device based on identifying optical distress signals. This device is applied to a rescue system and includes:

[0129] Acquisition unit 10 is used to acquire video images of light signals of suspected trapped persons captured by a mobile terminal;

[0130] The judgment unit 20, connected to the acquisition unit 10, is used to determine whether the suspected trapped person's optical signal video image is an optical distress signal video image after inputting the optical distress signal dynamic model into it.

[0131] As one specific implementation, the determination unit 20 includes:

[0132] The first extraction module is used to extract model features from the optical signal video images of suspected trapped persons to obtain model feature extraction data.

[0133] The preprocessing module, connected to the extraction module, is used to preprocess the model feature extraction data;

[0134] The second extraction module, connected to the preprocessing module, is used to input the preprocessed model feature extraction data into the preset light distress signal dynamic model and extract the feature vector of the light signal video image of the suspected trapped person.

[0135] The prediction module, connected to the second extraction module, is used to classify and predict the light signal video image of the trapped person based on the feature vector of the light signal video image of the trapped person, so as to determine whether the light signal video image of the suspected trapped person is a light distress signal video image.

[0136] The location determination unit 30, connected to the judgment unit 20, is used to determine the transmission location of the optical distress signal based on the video image of the optical distress signal and the location of the mobile terminal.

[0137] As one specific implementation, the position determination unit 30 includes:

[0138] The third extraction module is used to extract location image features from the video image of the light distress signal;

[0139] The first matching module, connected to the third extraction module, is used to match the extracted location image features with a publicly available or system cloud-based location database to obtain the possible transmission location of the light distress signal.

[0140] The second matching module, connected to the first matching module, is used to perform secondary matching on the possible transmission locations of the optical distress signal based on the location of the mobile terminal. After successful secondary matching, the transmission location of the optical distress signal is obtained.

[0141] in:

[0142] The location of the mobile terminal is determined based on the mobile terminal's positioning.

[0143] A successful secondary match indicates that the distance between the mobile terminal and the possible transmission location of the optical distress signal is within a preset range.

[0144] The alarm unit 40, connected to the location determination unit 30, is used to issue an alarm by displaying the video image of the light distress signal and the transmission location of the light distress signal, so that professionals can rescue the trapped person, thereby realizing rescue based on video dynamic analysis and identification of distress signals.

[0145] In one specific implementation, the device further includes a construction unit for constructing a dynamic model of the optical distress signal. The construction unit includes:

[0146] The collection module is used to collect images of various possible historical distress signals;

[0147] The fourth extraction module, connected to the collection module, is used to extract key image features corresponding to historical distress signal images;

[0148] Key image features include numerical features of image brightness, numerical features of image color, numerical features of image texture, and numerical features of image shape.

[0149] The input module, connected to the fourth extraction module, is used to input historical distress signal images and their corresponding key image features into the initial machine learning algorithm to obtain the initial dynamic model of the light distress signal.

[0150] The tuning module, connected to the input module, is used to train the initial dynamic model of the optical distress signal using training data and to tune the initial dynamic model of the optical distress signal using test data, thereby obtaining the dynamic model of the optical distress signal.

[0151] Example 3:

[0152] This embodiment provides a rescue system, which includes a mobile terminal and a server. The mobile terminal and the server are connected via a network. The mobile terminal is one of a smartphone, tablet computer, or laptop computer.

[0153] The mobile terminal is used to capture video images of light signals from suspected trapped individuals and transmit these images to the server.

[0154] The server is used to input the suspected trapped person optical signal video image into the optical distress signal dynamic model after receiving the suspected trapped person optical signal video image transmitted by the mobile terminal, and determine whether the suspected trapped person optical signal video image is an optical distress signal video image. If it is determined to be an optical distress signal video image, the server determines the transmission location of the optical distress signal based on the optical distress signal video image and the position of the mobile terminal.

[0155] The server is also used to issue alarms by displaying video images of the light distress signal and the location of the light distress signal, thereby enabling rescue based on dynamic video analysis and identification of distress signals.

[0156] The mobile terminal is used to capture optical signal video images of suspected trapped individuals and transmit them to the server. After receiving these video images, the server inputs them into the dynamic model of optical distress signals for analysis.

[0157] The server will determine whether these video images are distress signal video images, and determine the transmission location of the distress signal based on the distress signal video images and the location of the mobile terminal.

[0158] Once the location of the distress signal is determined, the server will issue an alarm by displaying the video image of the distress signal and its location, enabling rescue based on dynamic video analysis and identification of the distress signal.

[0159] This system can assist rescue operations by using video images of light signals. By analyzing and judging the signals on the server side, it can quickly identify the light distress signals and determine their transmission location, thereby providing guidance and support for rescue efforts.

[0160] Example 4:

[0161] This embodiment provides an electronic device, which includes a memory and a processor. The memory stores a computer program. When the processor runs the computer program stored in the memory, the processor executes the rescue method based on identifying light distress signals as described in Embodiment 1.

[0162] It is understood that the above embodiments are merely exemplary implementations used to illustrate the principles of the present invention, and the present invention is not limited thereto. For those skilled in the art, various modifications and improvements can be made without departing from the spirit and essence of the present invention, and these modifications and improvements are also considered to be within the scope of protection of the present invention.

Claims

1. A rescue method based on identifying optical distress signals, applied to a rescue system, characterized in that, The method includes the following steps: Step S1: Acquire video images of light signals from suspected trapped individuals captured by the mobile terminal; Step S2: After inputting the suspected trapped person's optical signal video image into the optical distress signal dynamic model, determine whether the suspected trapped person's optical signal video image is an optical distress signal video image: the method of using the recognition model to identify optical signal video images includes fixed international standard frequency lighting, or setting the rules; If the video image of the suspected trapped person is not a video image of a distress signal, repeat steps S1 to S2; if the video image of the suspected trapped person is a video image of a distress signal, determine the transmission location of the distress signal based on the video image of the distress signal and the location of the mobile terminal, and proceed to step S3. The step of determining the transmission location of the optical distress signal based on the video image of the optical distress signal and the location of the mobile terminal specifically includes: extracting location image features from the optical distress signal video image; matching the extracted location image features with a publicly available or system cloud-based location database to obtain the possible transmission location of the optical distress signal; performing a secondary matching based on the location of the mobile terminal to the possible transmission location of the optical distress signal, and obtaining the transmission location of the optical distress signal after a successful secondary matching; the location of the mobile terminal is determined based on the positioning of the mobile terminal, and a successful secondary matching means that the distance between the mobile terminal and the possible transmission location of the optical distress signal is within a preset range; Step S3: Issue an alarm by displaying the video image of the light distress signal and the location where the light distress signal was transmitted, thereby enabling rescue based on dynamic video analysis and identification of distress signals.

2. The rescue method based on identifying optical distress signals according to claim 1, characterized in that, In step S2, the suspected trapped person's optical signal video image is input into the optical distress signal dynamic model to determine whether the suspected trapped person's optical signal video image is indeed an optical distress signal video image. This specifically includes the following steps: Step A1: Extract model features from the optical signal video images of the suspected trapped person to obtain model feature extraction data; Step A2: Preprocess the model feature extraction data; Step A3: Input the preprocessed model feature extraction data into the preset optical distress signal dynamic model to extract the feature vector of the optical signal video image of the suspected trapped person; Step A4: Based on the feature vector of the trapped person's light signal video image, classify and predict the trapped person's light signal video image to determine whether the suspected trapped person's light signal video image is a light distress signal video image.

3. The rescue method based on identifying optical distress signals according to claim 1 or 2, characterized in that, Before step S1, there is also a step S0, which includes: constructing a dynamic model of the optical distress signal, which includes the following steps: Step S01: Collect images of various possible historical distress signals; Step S02: Extract key image features corresponding to historical distress signal images; The key image features include image brightness numerical features, image color numerical features, image texture numerical features, and image shape numerical features; Step S03: Input the historical distress signal images and their corresponding key image features into the initial machine learning algorithm to obtain the initial dynamic model of the optical distress signal; Step S04: Train the initial optical distress signal dynamic model using training data, and fine-tune the initial optical distress signal dynamic model using test data to obtain the optical distress signal dynamic model.

4. A rescue device based on recognizing optical distress signals, applied in a rescue system, characterized in that, include: The acquisition unit is used to acquire video images of light signals from suspected trapped individuals captured by a mobile terminal. The judgment unit, connected to the acquisition unit, is used to determine whether the suspected trapped person's optical signal video image is an optical distress signal video image after inputting it into the optical distress signal dynamic model. Methods of using recognition models to process optical signal video images include fixed international standard frequency lighting, or setting rules; A location determination unit, connected to the judgment unit, is used to determine the transmission location of the light distress signal based on the video image of the light distress signal and the location of the mobile terminal; The step of determining the transmission location of the optical distress signal based on the video image of the optical distress signal and the location of the mobile terminal specifically includes: extracting location image features from the optical distress signal video image; matching the extracted location image features with a publicly available or system cloud-based location database to obtain the possible transmission location of the optical distress signal; performing a secondary matching based on the location of the mobile terminal to the possible transmission location of the optical distress signal, and obtaining the transmission location of the optical distress signal after a successful secondary matching; the location of the mobile terminal is determined based on the positioning of the mobile terminal, and a successful secondary matching means that the distance between the mobile terminal and the possible transmission location of the optical distress signal is within a preset range; An alarm unit, connected to the location determination unit, is used to issue an alarm by displaying the video image of the light distress signal and the transmission location of the light distress signal, thereby enabling rescue based on dynamic video analysis and identification of distress signals.

5. The rescue device based on identifying optical distress signals according to claim 4, characterized in that, The determination unit includes: The first extraction module is used to extract model features from the optical signal video images of suspected trapped persons to obtain model feature extraction data. A preprocessing module, connected to the extraction module, is used to preprocess the model feature extraction data; The second extraction module is connected to the preprocessing module and is used to input the preprocessed model feature extraction data into the preset light distress signal dynamic model to extract the feature vector of the suspected trapped person's light signal video image. The prediction module, connected to the second extraction module, is used to classify and predict the light signal video image of the trapped person based on the feature vector of the light signal video image of the trapped person, so as to determine whether the suspected light signal video image of the trapped person is a light distress signal video image.

6. The rescue device based on identifying optical distress signals according to claim 4 or 5, characterized in that, The device further includes a construction unit for constructing a dynamic model of an optical distress signal, the construction unit comprising: The collection module is used to collect images of various possible historical distress signals; The fourth extraction module, connected to the collection module, is used to extract key image features corresponding to historical distress signal images; The key image features include image brightness numerical features, image color numerical features, image texture numerical features, and image shape numerical features; The input module, connected to the fourth extraction module, is used to input historical distress signal images and their corresponding key image features into the initial machine learning algorithm to obtain an initial dynamic model of the light distress signal. The optimization module, connected to the input module, is used to train the initial optical distress signal dynamic model using training data and optimize the initial optical distress signal dynamic model using test data to obtain the optical distress signal dynamic model.

7. A rescue system, characterized in that, It includes a mobile terminal and a server, which are connected via a network. The mobile terminal is used to capture video images of light signals from suspected trapped individuals and transmit these images to the server. The server is used to input the suspected trapped person optical signal video image into the optical distress signal dynamic model after receiving the suspected trapped person optical signal video image transmitted by the mobile terminal, and determine whether the suspected trapped person optical signal video image is an optical distress signal video image. If it is determined to be an optical distress signal video image, the server determines the transmission position of the optical distress signal based on the optical distress signal video image and the position of the mobile terminal. The method of using a recognition model to analyze optical signal video images includes fixing the frequency of illumination according to international rules, or setting rules; determining the transmission location of the optical distress signal based on the optical distress signal video image and the location of the mobile terminal specifically includes: extracting location image features from the optical distress signal video image; matching the extracted location image features with a publicly available or system cloud-based location database to obtain the possible transmission location of the optical distress signal; performing a secondary matching based on the location of the mobile terminal on the possible transmission location of the optical distress signal, and obtaining the transmission location of the optical distress signal after a successful secondary matching; the location of the mobile terminal is determined based on the mobile terminal's positioning, and a successful secondary matching means that the distance between the mobile terminal and the possible transmission location of the optical distress signal is within a preset range; The server is also used to issue alarms by displaying video images of the light distress signal and the location of the light distress signal, thereby enabling rescue based on dynamic video analysis and identification of distress signals.

8. An electronic device comprising a memory and a processor, characterized in that, The memory stores a computer program, and the processor is configured to run the computer program to implement the rescue method based on the identification of light distress signals as described in any one of claims 1-3.

Images