46 results about "Flame" patented technology

Reigniting a flame in a volatile organic compound (VOC) detector in the event that the flame has gone out. In one implementation, a signal is received at a handheld personal computer indicating that a flame in the VOC detector has gone out. The flame in the VOC detector may then be reignited using the handheld personal computer and a Bluetooth enabled device facilitating communication between the handheld personal computer and the VOC detector.

The invention discloses a security and protection scene flame detection method based on deep learning, which belongs to the technical field of security and protection, and mainly comprises the following steps of: detecting a suspected flame area of a picture decoded by a monitoring video through a single-stage detection model which is trained based on a neural network and is used for identifying aflame shape; according to the identified suspected flame area, performing video frame interception on the area corresponding to the suspected flame area in the video to obtain a video frame; dividingthe video frame into N sub-segments, and sampling one frame from each sub-segment to obtain sampling truth; and finally, inputting the sampling frame into a class behavior identification classification model which is trained based on a neural network and is used for identifying the dynamic change of the flame so as to classify whether the flame is the flame. According to the method, the appearance features of the single frame of the suspected flame are extracted through the single-stage detection model, the dynamic information of the front frame and the rear frame is considered, the final classification effect is greatly improved through richer features, and the real-time performance and high efficiency of flame detection are improved.

The invention discloses a multi-factor flame recognition method suitable for an embedded platform. The method comprises the steps of building a fire sample library, wherein the fire sample library isfrom a network fire picture and a combustion experiment picture; obtaining a plurality of live video frames separately; extracting moving objects in each live video frame separately and obtaining oneor more quasi-fire areas; carrying out fire confirmation on one or more quasi-fire areas according to the fire sample library and judging whether fire information exists in the one or more quasi-fireareas or not; and if so, carrying out fire classification on the fire information and generating corresponding alarm signals according to the classified fire classes. According to the multi-factor flame recognition method, the obtained live video frames are processed separately, whether the fire information which can be developed into a fire appears or not in a live monitoring environment is analyzed in real time, fire confirmation is carried out again, whether the fire information really exists or not is further judged, fire class analysis is also carried out on the fire information and alarms are generated, so that fire information recognition is more accurate.

The invention discloses a flame detection method based on an improved RetinaNet network, and the method comprises the steps: S1, collecting N pictures with flame pictures as a training data set, and marking flames in the training data set; S2, a SandGlass module being used for replacing the residual error module, so that an improved RetinaNet network is obtained, and the improved RetinaNet network is recorded as SG-ResNet 50, wherein the SandGlass module comprises a first depth separable convolution, a first convolution, a second convolution and a second depth separable convolution which are connected in sequence; S3, constructing a feature pyramid network, and adding a segmentation branch behind each layer of features output by the feature pyramid network; S4, training the constructed improved RetinaNet network to obtain a trained flame detection model; and S5, carrying out flame detection on the obtained video by adopting the flame detection model obtained in the step S4. According to the invention, the SandGlass module is used for replacing a residual error module of the existing RetinaNet network, so that the flame detection speed is improved; segmented supervision signals are provided by utilizing the color characteristics of flame, and the flame detection precision is improved.

The invention discloses a real-time flame monitoring system and method based on an Internet of Things distributed architecture, and belongs to the technical field of fire-fighting Internet of Things and artificial intelligence algorithm research and application. The system comprises a plurality of intranet subsystems, a data processing center and a client. Intelligent identification of a flame target in the current monitoring environment can be realized, automatic alarm can be achieved in the early stage of fire, and the fire situation is visualized; when a fire is detected, alarm signals including text information and acousto-optic information can be generated immediately, fire alarm information and fire scene pictures are uploaded to a cloud server, the cloud server pushes the alarm information to a mobile phone terminal of a user, and the mobile phone terminal reminds the user in a ringing mode after receiving the pushed alarm information; and a deep learning algorithm realizes early-stage fire alarm processing, the reliability and instantaneity of fire alarm are improved, and s the fire-fighting management cost is reduced by through cloud remote communication.

The invention discloses an interaction control method and system for simulating flame. The interaction control system comprises flame simulation equipment and fire-extinguishing equipment. The flame simulation equipment comprises a first induction module, a first control module, a second control module and flame simulation generation equipment; and a signal emission module is arranged on the fire-extinguishing equipment. The interaction control method comprises the following steps that the signal emission module emits a first induction signal, and the first induction module is used for receiving the first induction signal; and the first control module judges whether the first induction signal received by the first induction module meets the induction signal type corresponding to a currentsimulated flame type or not according to a corresponding relationship of the simulated flame type and the induction signal type, and if yes, the second control module adjusts the simulated flame generated under the current simulated flame type. According to the interaction control method and system for simulating the flame, through the signal interaction of a fire extinguisher and the flame simulation equipment, the actual simulation of a fire-extinguishing scene is realized, and the requirement of fire safety training is effectively met.

The invention discloses a cloud-side collaborative power distribution station flame intelligent monitoring method and system. The method comprises the following steps: collecting a flame data set and marking the flame data set; carrying out migration training and then quantifying into a model Qint8; initializing the summation tree to obtain a video image; filtering out the flame redundant frame to obtain a flame key frame pkey; inputting the flame key frame pkey into a detection network Qint8 for detection; extracting m samples, reminding a user to mark, and uploading the samples marked by the user; enabling the cloud platform to randomly combine the received samples and the existing samples again according to a certain proportion to obtain a new flame data set; then carrying out migration training in combination with an existing network weight; after training is completed, updating the training weight in a soft updating mode; and quantifying into a model Qint8 and issuing. The system comprises a cloud platform, an edge computing platform and video equipment. Through a mode of searching data at an edge end and training the data at a cloud end, the utilization rate of computing resources is improved, task delay is reduced, and the detection precision can be continuously improved.

The invention is suitable for the technical field of fire monitoring, and provides a method for preventing flame identification false alarm, which comprises the following steps: when a suspected flame area in a monitoring range is detected, acquiring a plurality of continuous frame images; according to the contour of the suspected flame area in the frame image, respectively constructing corresponding flame identification frames; comparing the flame recognition frames in the plurality of frame images to obtain change information among the flame recognition frames; and judging whether the change information accords with preset flame characteristic information or not, and if yes, generating flame early warning and/or fire alarm. A system for preventing flame recognition misinformation, a storage medium for storing a computer program for executing the method, and a computer device for implementing the method are also provided. Therefore, according to the invention, the false alarm problem can be greatly reduced, and the working efficiency of monitoring personnel is improved.

The invention discloses a flame detection method based on an improved ViBe algorithm and a lightweight convolutional network. The flame detection method comprises the following steps: carrying out flame foreground detection by utilizing the improved ViBe foreground detection algorithm; and further judging whether the suspected flame area detected by the ViBe foreground detection algorithm is a flame object or not by utilizing the constructed lightweight convolutional neural network, and constructing the lightweight flame detection convolutional neural network based on the SE attention mechanism. According to the invention, a ViBe foreground detection algorithm is used for detecting a suspected flame foreground target, eliminating dynamic background noise points and removing an interferenceobject which is similar to flame in color and does not have flicker characteristics; compared with an existing target detection deep convolutional network, the method has the advantages that the flame detection speed is increased; an SE attention mechanism is embedded into the lightweight flame detection convolutional network, so that the flame detection speed is ensured while the flame detectionprecision is improved.

The invention relates to a substation fire detection method and device based on terminal power business edge calculation, which are applied to an intelligent gateway, and the intelligent gateway comprises an edge computing unit. The method comprises the following steps: receiving a transformer substation monitoring image sent by image acquisition equipment; the transformer substation monitoring image is an image obtained by shooting a target transformer substation in real time by the image acquisition equipment; sending the transformer substation monitoring image to the edge calculation unit; inputting the transformer substation monitoring image into a pre-trained flame detection model through the edge calculation unit to obtain a flame detection result; the flame detection result is a result obtained by labeling a flame region in the transformer substation monitoring image by the pre-trained flame detection model; and generating a fire detection result for the target transformer substation based on the flame detection result. The method can improve the fire inspection efficiency.

The invention relates to a processing method and device for simulating an explosion flame effect in a GIS system based on graphics. The method comprises the steps that entities and expressions are abstracted into geometries and materials in advance; the explosion special effect is compiled for materials, and codes of a vertex shader and a fragment shader are directly compiled by utilizing self-defined coloring material materials to define the materials; taking an initial vertex of a three-dimensional sphere geometry as basic data of a flame explosion special effect; obtaining normal vectors ofvertexes of the three-dimensional spherical surface needing to simulate the explosion flame through a vertex shader; obtaining a current time and a position coordinate position; calculating the latest position of the current vertex through a Berlin noise algorithm; and loading the flame special effect chartlet by using a fragment shader to finish overall special effect rendering. According to theinvention, the special effect of flame explosion can be conveniently realized, the explosion flame effect can be simulated in a GIS system, the user operation and use are convenient, the realizationis simple, and the reliability is high.

Flame detection serves as an important link of fire prevention and control, and has high requirements for real-time performance, anti-interference performance and accuracy. The flame target detection method at the present stage lacks comprehensive research on the three indexes, and in order to solve the problem, the invention provides the light-weight anti-interference flame detection method based on the improved YOLOv4-tiny. A flame detection model of a double-flow structure is designed by utilizing the dynamic characteristic that flame changes along with time. The method comprises the following steps: firstly, carrying out lightweight improvement on a backbone network of YOLOv4-tiny by adopting depth separable convolution; secondly, in the feature extraction stage, the learning ability of the network for shallow features is improved by further fusing multi-scale features, and meanwhile, an ECA channel attention module is introduced into the FPN, so that the precision is further improved; and finally, an IOU (Intersection over Union) post-processing algorithm is adopted to effectively shield the interference of the fire-like target. In a dataset aspect, an own flame detection dataset is created. Experiments prove that the accuracy, the anti-interference performance and the detection time of the method are comprehensively improved.

The invention discloses an open fire detection method and system based on flame runout modeling. The method comprises the steps of acquiring a video image; preprocessing continuous frame images in thevideo image; carrying out labeling and edge detection on the image to obtain an edge detection result of the image, and selecting a closed region with the largest area from the edge detection resultto serve as a target region; inputting the target area into a pre-trained flame runout detection model; analyzing the target area by using a flame runout detection model, and determining whether openfire exists in the image or not;, wherein the flame runout detection model is obtained by training a plurality of groups of data.

According to the flame detection method and system and the computer readable storage medium provided by the invention, the suspected flame region is re-detected on the premise of constructing the flame tripartite graph and removing irrelevant background interference on the to-be-detected image, so that the flame detection accuracy is improved, the requirement on the flame detection performance of the network model is reduced, and the flame detection efficiency is improved. According to the method, fire detection can be carried out through the lightweight neural network, so that the method is suitable for an edge device end, a suspected flame area is identified, and positioning of a fire point is realized.

The invention discloses an induction control method and system for a simulated flame. The system comprises a flame simulation device and a fire extinguishing device, wherein the flame simulation device comprises a signal induction module, a first control module, a second control module and a simulated flame generation module; and the fire extinguishing device is provided with a signal transmittingmodule. The method comprises the following steps that: the signal transmitting module transmits a first induction signal, and the signal induction module is used for receiving the first induction signal; and the first control module judges whether the first induction signal received by the signal induction module conforms to a corresponding induction signal type under the current simulated flametype according to a corresponding relation between the simulated flame type and the induction signal type, and if so, the simulated flame generated under the current simulated flame type is adjusted through the second control module. According to the scheme, real simulation for a fire extinguishing scene is realized through signal interaction between a fire extinguisher and the flame simulation device, and requirements of fire safety training are effectively met.

The invention relates to emergency command management, in particular to an emergency command management system based on the Internet of Things, which comprises a server and a distributed image acquisition module connected with the server, the server constructs a flame detection model for detecting whether a flame target exists in an image acquired by the image acquisition module through the flame detection model construction module, and records the detection target obtained by the flame detection model through the detection target recording module; a first detection target judgment module is used for carrying out preliminary authenticity judgment on the recorded detection targets, a server classifies the detection targets passing the preliminary authenticity judgment through a detection target classification module, and a second detection target judgment module is used for carrying out secondary authenticity judgment on the classified detection targets. The server counts flame targets one by one by using a flame target counting module; according to the technical scheme provided by the invention, the defects that fire identification cannot be quickly and accurately carried out and an optimal escape path cannot be provided for escape crowds can be effectively overcome.

The invention discloses a water-cooled wall leakage monitoring system and method based on an inverse Abel transform algorithm, relates to the field of equipment monitoring, and relates to a water-cooled wall leakage monitoring method and a water-cooled wall leakage monitoring system. The method comprises the following steps: acquiring a flame image, transmitting the acquired image to a server through a communication cable for flame analysis and processing, then obtaining the development condition of water-cooled wall leakage according to a flame analysis result, and finally collecting and storing the acquired data. According to the method, the leakage degree is reversely analyzed through the influence of leakage on flames, and the accuracy and uniformity of cold wall leakage analysis are guaranteed; two indexes are adopted to measure the leakage degree of the water cooling wall, predictive analysis of leakage in different stages can be carried out, and criteria are provided for operators; in addition, manual fire observation operation after traditional leakage is avoided, and the risk operation during manual fire observation is effectively reduced.

The embodiment of the invention discloses a lag function searching method and device, electronic equipment and a storage medium. The method the following steps: acquiring each function contained in a lagging flame diagram and the cumulative sampling frequency of each function in the lagging flame diagram; acquiring a normal flame map at a historical moment; for each function included in the lagging flame map, comparing the cumulative sampling frequency of the functions included in the lagging flame map with the cumulative sampling frequency of the same functions in the normal flame map to obtain a comparison result; according to a comparison result, screening out a preliminary screening function from all functions contained in the lagging flame map; and selecting K lag functions from the preliminary screening function according to the type of the function and the type priority. A preliminary screening function can be screened from a lagging flame map; and then, the K lag functions are selected from the preliminary screening function according to the types of the functions and the type priorities, so that the problem of relatively high cost of screening and analyzing work of the flame diagram in the prior art is improved.

The invention discloses a motion control card device for a flame cutting numerical control system. The product comprises components an upper computer IPC, which is connected with a lower computer through a peripheral component expansion interface PCI, wherein the lower computer comprises a motion control card, the motion control card is provided with PULSE/DRI, ECA/ECB/ECZ and LMT+/LMT-, the PULSE/DIR is connected with an X/Y/Z-axis servo motor driver through a pulse/direction, the X/Y/Z-axis servo motor driver is connected with the ECA/ECB/ECZ in a limited mode through an encoder, the X/Y/Z-axis servo motor driver is connected with a travel switch, and the travel switch is connected with the LMT+/LMT-. The motion control card device is used for the flame cutting numerical control system.

The invention discloses a flame target detection method, electronic equipment and a storage medium. The method comprises the following steps: S1, obtaining an image; s2, training a flame detection model by using a deep learning convolutional neural network to obtain a detection model yo-v5; s3, each frame of the image is detected through the detection model yo-v5, and a preliminary detection result is obtained; s4, filtering the preliminary detection result through a filtering module based on flame image features; and S5, judging that the detection target meeting the judgment condition is flame, and otherwise, judging that the detection target is misjudged. According to the invention, the false detection rate during flame target detection can be reduced.

The invention discloses a flame identification method and system, and a medium. The method comprises the following steps: S100, obtaining a flame image; S200, performing image processing through a flame edge repairing algorithm to generate a data set; S300, inputting the processed data set for training into a YOLOV5 network for training to obtain a flame state detection model; and S400, inputting the obtained actual data set into the flame detection model, and outputting a flame state detection result. With the adoption of the system, a blast furnace operator does not need to guess and deduce by experience to operate, but directly observe the real scene of the specific condition in the boiler, and can ensure the normal operation of the boiler by taking corresponding measures according to the actual combustion condition, so that the boiler achieves safe production, stable production and high yield, and an enterprise acquires huge economic benefits.

The invention discloses a multi-feature information fusion flame detection system and method. The system comprises an infrared camera, a video monitoring camera, a communication cable, a support, a monitoring area and a server computer terminal. The infrared camera is connected with the server computer terminal through a communication cable; the video monitoring camera is connected with the server computer terminal through a communication cable; the infrared camera and the video camera are fixed through a support and form an overlapped monitoring area; and the server computer end performs fusion analysis according to the images transmitted by the infrared camera and the video monitoring camera to judge whether flame exists or not. The flame detection system is low in manufacturing cost, good in detection effect, high in intelligent degree and capable of effectively and accurately detecting the fire.

The invention discloses a data acquisition card device for a flame cutting numerical control system. The product comprises an upper computer IPC, and is characterized in that the upper computer IPC isconnected with a lower computer through a peripheral component expansion interface PCI, the lower computer comprises a multifunctional data acquisition card, the multifunctional data acquisition cardis provided with an I/O and an AT, the I/O is connected with a switch, a relay and a control panel through a DI/O, and the AT is connected with a laser displacement sensor and a pressure sensor through analog input quantity. The data acquisition card device is used for a flame cutting numerical control system.

The invention relates to the technical field of fire retardance safety, and discloses a fire retardance performance evaluation method and system of a flame arrester, and electronic equipment, and the method comprises the following steps: carrying out the appearance reliability evaluation of the flame arrester; evaluating the reliability of the internal structure of the flame arrester; fire arresting key parameter indexes of fire arresting units of the fire arresters are detected and evaluated; and based on the evaluation results of the flame arrester appearance reliability evaluation, the flame arrester flame retardance performance reliability evaluation and the flame arrester flame retardance unit flame retardance key parameter index detection evaluation, performing flame arrester flame retardance performance reliability evaluation. According to the technical scheme provided by the invention, a non-destructive method is adopted to test and evaluate the flame retardance of the flame arrester, so that the test efficiency is high, the test time is short, and stable operation of the device is ensured.