452results about How to "Reduce false detection rate" patented technology

The invention provides a semiconductor chip defect automatic detection method and system based on hyperspectral imaging, and relates to the technical field of chip detection, and the method comprises the steps: collecting a chip hyperspectral image, carrying out the image enhancement and illumination non-uniformity correction through an improved neuromorphic adaptive processing network and a dual-path diffusion model, and carrying out the detection of the hyperspectral image. Extracting multi-level features by using a dynamic mixed feature enhancement network based on neural architecture search, fusing the features through a cross-layer attention mechanism, generating an initial defect detection result by using a multi-branch expert decision-making module and an uncertainty dynamic integration strategy, constructing a space-time knowledge inference graph, and determining a defect detection result; and performing reliability evaluation and optimization on the initial defect detection result in combination with the form similarity, the spatial position relationship and the process parameter influence model to obtain a final defect detection result.

The invention relates to the technical field of electrical digital data processing, in particular to an anomaly detection and processing method and system based on deep learning. The method comprises the following steps: firstly, performing feature extraction and fusion on a log text sequence, a log timestamp sequence and a running state index sequence set to obtain an initial feature vector sequence; then, constructing a hierarchical graph neural network to perform feature updating, propagation and fusion on the initial feature vector sequence to obtain a multi-level fusion feature vector sequence, and calculating a standardized mode vector sequence in combination with a log text sequence and a running state index sequence set; obtaining a self-adaptive abnormal threshold value according to the standardized mode vector sequence and detecting candidate abnormal logs; and finally, verifying the candidate abnormal logs, determining verified abnormal logs, classifying the verified abnormal logs, generating a processing suggestion text according to a classification result, and carrying out early warning prompt. The false detection rate can be reduced, and the log anomaly detection accuracy is improved.

The invention discloses a PCB production line defect detection analysis method and system based on an identification model, particularly relates to the technical field of PCB quality detection, and is used for solving the problems of low PCB defect detection precision and high omission ratio in the prior art. The method comprises the following steps: scanning the whole surface of a PCB to generate a thickness distribution diagram, identifying a thickness non-uniform region by adopting a threshold segmentation algorithm, and marking a key detection region; potential defects irrelevant to process fluctuation are identified through historical production data analysis; and adjusting the polarization angle of the light source to analyze the reflectivity difference, and identifying the hidden defect area in combination with the microstructure of the material. The union set of the potential defect area and the hidden defect area is marked as the suspected incomplete area, and the influence of the suspected incomplete area on the performance of the PCB is analyzed according to the quality standard, so that the detection precision and comprehensiveness are improved, and the high-quality control of the PCB is realized.

The invention discloses an intersection blind area auxiliary driving method and system based on V2X, and relates to the technical field of intelligent auxiliary driving, and the method comprises the steps: collecting multi-modal data through a vehicle-mounted sensor, carrying out the preprocessing of the multi-modal data, carrying out the feature extraction of the preprocessed multi-modal data, and forming a modal feature vector; based on the multi-modal feature vector, a deep learning model is adopted for comprehensive perception, and the position of an obstacle in the blind area of the intersection is recognized; monitoring the position change of the obstacle in the blind area of the intersection in real time, and predicting the future movement track of the obstacle; evaluating the risk level of the vehicle passing through the intersection according to the future movement track of the obstacle; and receiving real-time traffic information through V2X communication, and generating a driving decision in combination with the real-time traffic information and the risk level. According to the invention, a safer and more efficient driving decision is finally generated through the combination of real-time obstacle position change monitoring, motion trajectory prediction, risk assessment and V2X communication.

The invention relates to the technical field of braid quality detection in industrial production, in particular to a defect data processing query method and system for braid visual inspection. The method comprises the steps of obtaining braid image data, performing enhancement processing on an image based on a dynamic illumination compensation algorithm, and generating a high-quality image in combination with material characteristics; inputting the enhanced image data into a deep learning defect detection model, detecting and classifying defect areas in the image, and updating the model weight based on defect feature parameters; performing multi-dimensional indexing processing on the detected defect feature data to generate a defect distribution database supporting real-time query; and the defect distribution database and the production process parameter sequence are combined, the causal relationship of defect occurrence is analyzed, process optimization parameters are generated and fed back to the detection module, and self-adaptive updating of model parameters is completed. According to the invention, high precision, high efficiency and closed-loop optimization of braid defect detection are realized, and intelligent support is provided for industrial production quality control.

The invention relates to the technical field of assembly welding, and particularly discloses an assembly welding positioning method and device for automobile parts. A welding state image between a target hush pipe and a target partition plate is collected through a camera; carrying out welding state mask segmentation on the welding state image by using an image identification and analysis algorithm based on computer vision, and then carrying out welding state feature extraction on the segmented welding detection area state mask image; and according to the expression obtained through neighborhood importance optimization of the extracted welding detection state characteristics, intelligent detection on whether defects exist in the welding state between the target hush pipe and the target partition plate or not is achieved. In this way, tiny defects such as tiny cracks and internal pores can be captured, and therefore the accuracy of defect recognition is improved. Moreover, the detection speed is greatly improved, the consistency and stability can be maintained, and the probability of missing detection or false detection is reduced, so that each welding point is ensured to be subjected to strict quality control.

The invention discloses a power transmission line unmanned aerial vehicle inspection picture defect identification method and system. The method comprises the following steps: carrying out picture resolution amplification processing on a power transmission line unmanned aerial vehicle inspection picture; inputting the processed picture into a large target component identification model, identifying large target component defects and large target components where small target component defects are located, obtaining confidence coefficients corresponding to the identified large target component defects, and retaining each identified large target component defect greater than a confidence coefficient threshold value; cutting an image of the large target part where the identified small target part defects are located, inputting the image into a small target part identification model, identifying the small target part defects, obtaining confidence coefficients corresponding to the small target part defects, and retaining each identified small target part defect greater than a confidence coefficient threshold value; and summarizing the recognized defects of the large target part and the recognized defects of the small target part, and outputting a defect recognition result. According to the method and the device, the problem of defect identification of images with high resolution and concentrated defect distribution can be solved.

The invention discloses a photovoltaic cell hot spot intelligent detection method fusing multi-modal data. The method comprises the following steps: S1, collecting multi-modal data of a photovoltaic cell; s2, carrying out preprocessing on the multi-modal data; s3, constructing a photovoltaic cell sample data set; s4, extracting features of the multi-modal data; s5, fusing the multi-modal features to generate a comprehensive feature vector; s6, training and optimizing a dynamic hot spot detection model based on the space-time convolutional network; s7, identifying a hot spot area in the photovoltaic cell by using the model; s8, positioning and outputting a hot spot detection result; and S9, optimizing the detection model in real time in combination with a dynamic adaptive algorithm. According to the method, the multi-modal data fusion, the space-time convolutional network and the dynamic adaptive optimization algorithm are introduced, so that the hot spots in the photovoltaic cell can be detected and positioned more accurately and quickly, the false detection rate in a complex environment is effectively reduced, and the operation efficiency and reliability of a photovoltaic cell system are remarkably improved.

The invention discloses a road small target detection algorithm based on Real-ESRGAN and improved YOLOv5, and the algorithm comprises the steps: S1, preparing a data set, collecting data, and dividing the data into a training set, a verification set and a test set; s2, building a YOLOV5 algorithm model, inputting data and training the data to obtain a YOLOV5 algorithm model weight; the trained YOLOV5 algorithm model weight is used to detect whether a small road target appears, and if yes, a detected image frame is intercepted and stored; s3, performing super-resolution reconstruction on the detected image frame, and outputting a detection frame image after the super-resolution reconstruction; and S4, identifying the output detection frame image after super-resolution reconstruction, and updating a detection result after classification. By adopting the Real-ESRGAN and the improved YOLOv5 detection algorithm, the detection confidence and accuracy of the small road target are improved, and the false detection and missing detection conditions are reduced.

The invention provides a forestry pest monitoring method and system based on machine learning, and the method comprises the steps: carrying out the principal component analysis of a forestry pest image, obtaining a plurality of recognition indexes of the pest of a target forestry region, determining an effective feature vector when the image recognition of the forestry pest is carried out according to all recognition indexes, and carrying out the recognition of the effective feature vector. Determining a first-level correlation feature of the pest image in the target forestry region through the effective feature vector; extracting a plurality of environmental influence factors from the monitoring image, and determining secondary correlation features of the pest image in the target forestry area according to all the environmental influence factors; carrying out interactive fusion on the first-level correlation features and the second-level correlation features to obtain multi-level distribution features for carrying out image recognition on the diseases and insect pests; and identifying the current pest and disease distribution of the target forestry area based on the multi-level distribution characteristics through a pest and disease identification model. By adopting the scheme of the application, diversified monitoring of the forestry diseases and insect pests can be realized under a complex background, so that the false detection rate of the forestry diseases and insect pests is reduced.

The invention discloses a point cloud image data fusion method, system and device based on an improved DBSCAN clustering algorithm and a storage medium, and belongs to the field of multi-sensor data fusion, the method uses an acquisition system composed of a laser radar and a camera to complete point cloud and image shooting and acquisition of a detected target scene, and the point cloud and image shooting and acquisition have different characteristics; the sensor fusion mode is used for target detection, so that advantage complementation can be realized; an improved DBSCAN algorithm is used for completing detection target point cloud clustering and generating a three-dimensional bounding box, finally, a point cloud cluster bounding box model is projected to a corresponding image pixel position, fusion of image data and point cloud data is completed, the DBSCAN point cloud clustering algorithm is improved, automatic calculation of a minimum neighborhood and an initial radius value is achieved, and finally, point cloud detection is completed. And the efficiency of detecting the target point cloud clustering is improved, so that the robustness and the accuracy of the detected target are improved, and the omission ratio and the false detection rate of target detection under the condition that the illumination condition is insufficient are reduced.

The invention discloses an identification method and system based on multi-sensor information fusion. According to the method, firstly, a one-dimensional recognition framework and a two-dimensional recognition framework are constructed, and sensor detection results and reliability thereof are described respectively; grouping is carried out according to sensor feature information, calculation is simplified, and the information fusion difficulty is reduced. And then the reliability of the sensor is evaluated by adopting an AHP-TOPSIS method, and evidence is corrected. And finally, integrating each group of evidences by applying a D-S evidence theory fusion rule to generate a final fusion detection result. The method effectively fuses multi-sensor data, improves decision-making efficiency, and is suitable for intelligent identification tasks in a complex environment.

The invention discloses an industrial defect detection method based on reverse distillation learning, which relates to the technical field of industrial defect detection, and comprises the steps of generating and filtering a noise picture, and fusing reverse distillation learning with an attention mechanism. According to the method, the fine defect detection and positioning capability under the complex background is improved, and particularly, the defects of small defect area and low background contrast ratio under the condition of complex background texture or relatively high noise level are overcome; according to the performance, the image level AUROC is 99.0%, the pixel level AUROC is 98.23%, the AUPRO can reach 95.02%, and the AUROC and the AUPRO are far higher than other indexes based on distillation learning. In the aspect of real-time performance, the detection speed of a single picture is high and can reach 25ms, the real-time performance is in the front, and the defect detection performance and the defect positioning performance are higher; noise synthesis and filtering are used for improving the small defect detection rate and reducing the false detection rate; the MFF + attention mechanism is used for improving the logic defect detection rate.

The invention discloses a ground penetrating radar small target B-scan identification method and system for underground facilities of a transformer substation, and relates to the field of ground penetrating radar data processing. The problems of false detection and high omission ratio caused by a large number of small target samples in underground facility ground penetrating radar detection in a power grid and the problems that a traditional image processing method depends on manual identification and labeling, and the data processing accuracy and processing speed are difficult to improve are solved. The method comprises the following steps: scanning an underground facility along a motion direction according to a UWB ultra-wideband radar system to obtain a B-scan graph; identifying the B-scan image by adopting a LableImage, and obtaining a characteristic waveform in the B-scan image; a GPR-Bfnet recognition model is constructed, and the GPR-Bfnet recognition model is trained according to the training set; and the trained GPR-Bfnet identification model is used to carry out identification and classification on the characteristic waveform in the B-scan image. The method is applied to the fields of engineering, geological exploration and military.

The invention discloses a short message user abnormal behavior detection method, system and device for a short message user with an abnormal behavior and a storage medium, and the method comprises the steps: obtaining an original short message record, and constructing a complex network according to the original short message record; initial nodes of the complex network represent short message users, edges between the initial nodes represent that short message records exist between the initial nodes, a sub-network of each initial node is obtained according to the complex network, and statistical characteristics of the sub-networks are extracted; the statistical features comprise the actual node number and the actual edge number of the sub-network, calculating the deviation degree of the statistical features and the node feature model, and screening the initial nodes according to the deviation degree to obtain the short message user with the abnormal behavior. The method can reduce false detection probability, save learning time and calculation cost, and can be widely applied to the technical field of short message security detection.

The invention discloses a deep learning-based picture frame information intelligent identification method and system and a storage medium, and the method comprises the steps: extracting a picture frame range under a layout, and filtering the picture frame range to obtain a target picture frame range; on the basis of a graph recognition algorithm, identifying a picture tag bar cell in each target picture frame range, and predicting a whole picture tag bar area on the basis of a yov8 prediction model; and on the basis of the picture tag bar cells and the text bounding box, determining picture tag information in the target picture frame range. According to the method, graph and text information in the DWG file is automatically identified and analyzed through advanced graph and image processing and deep learning technologies, key data can be rapidly and accurately extracted from a large number of drawings, and the efficiency and quality of building design and construction are improved.

The invention discloses an industrial defect detection and positioning method and system based on frequency domain enhancement, relates to the field of computer vision and image processing, can discover abnormal modes in an industrial image from a plurality of perspectives, and also has a relatively high reasoning speed. The method comprises the following steps: processing an input image through a high-pass filter to obtain a high-frequency image; inputting the original image and the high-frequency image into an improved U-Net network integrated with a double-domain feature selection module; image reconstruction is carried out in a spatial domain and a frequency domain at the same time; an anomaly score and a localization map are calculated based on the reconstruction result. According to the invention, by combining spatial domain and frequency domain analysis, comprehensive characterization of local pixel change and global frequency mode of the image is realized, and the accuracy of anomaly detection and positioning is improved. The method achieves the optimal performance on a plurality of real industrial scene data sets. The method is suitable for abnormal detection and abnormal area positioning of industrial product images.

The invention relates to the field of ocean internal solitary wave automatic detection, and discloses an internal solitary wave detection method based on sentinel No.3 satellite altimeter data, which comprises the following steps: acquiring sentinel No.3 satellite altimeter data of a sea area to be detected, and preprocessing the data; sequentially extracting a posterior scattering coefficient of each coordinate point of the Ku wave band and the C wave band, and response characteristic parameters SSHA and SWH; sequentially calculating the difference mean square slopes of the posterior scattering coefficients of the Ku wave band and the C wave band; the response characteristic parameters are smoothed twice, and variables # imgabs0 # and # imgabs1 # are obtained; and respectively calculating a differential mean square slope and a local standard deviation of variables # imgabs2 # and # imgabs3 #, and extracting data fluctuation characteristics by setting a threshold value so as to realize internal solitary wave detection. The method disclosed by the invention has higher accuracy and lower false detection rate, and still has a good detection effect under the condition that the propagation direction of the inner solitary wave is greatly different from the running direction of the satellite.

The invention provides an image feature detection method based on flexible connection, relates to the technical field of image analysis and processing, and aims to solve the defects of an existing image feature detection method in the aspects of accuracy, processing speed, detection sensitivity, false detection rate and the like. The method comprises the following steps: firstly, acquiring an input image, converting the input image into a grayscale image, carrying out Gaussian filtering processing, then calculating a gradient magnitude and direction, and carrying out non-maximum suppression and dual-threshold processing to obtain a preliminary edge image and a final edge image; and then connecting the weak edge pixels, extracting image features through a connected region analysis algorithm, calculating parameters such as a geometric center, a perimeter and an area, judging whether the features meet requirements or not according to a preset rule, and finally outputting the image features meeting the requirements. According to the method, parameters can be flexibly adjusted, different images can be adapted, the method is excellent in the fields of object recognition, medical image analysis, industrial detection and the like, the detection accuracy and efficiency are improved, the false detection rate is reduced, and the method has wide application prospects and practical value.

The invention relates to a woolen sweater defect detection method. The method is characterized by comprising the following four main steps: a data preprocessing step, a model construction step, a model training step and a defect detection step. Wherein in the model construction step, strategies such as an attention mechanism and a network structure optimization are introduced, so that the identification capability of the model on woolen sweater defects is improved; in the detection process, automatic positioning and classification of defect areas are realized through a series of key functions. The method is high in detection efficiency, low in comprehensive detection cost and high in detection accuracy.

The invention discloses a pointer instrument detection and identification reading method based on improved YOLOv8n, and the method comprises the steps: obtaining the image data of a pointer instrument, and sequentially carrying out the data expansion and dial marking, and obtaining a first marking data set; performing optimization adjustment based on partial convolution modules in a backbone network and a neck network of the YOLOv8n network structure and a convolution module in a head network to obtain an improved YOLOv8n network model; training the improved YOLOv8n network model based on the first annotation data set to obtain a dial region detection model; inputting the first annotation data set into a dial region detection model to obtain a dial data set; performing secondary annotation based on the dial data set to obtain a second annotation data set, and training the dial area detection model to obtain a target frame detection model; inputting the to-be-detected instrument image into the target frame detection model to obtain a target frame; and obtaining the pointer reading of the to-be-measured instrument based on the target frame. And the reading identification efficiency and accuracy of the pointer type instrument are improved.

The invention discloses a target detection method, system and device based on improved YOLOv5s and a storage medium, and the method comprises the steps: providing a picture data set, and manually marking a detection target for a picture in the data set; constructing a YOLOv5s network model, adding a GAM attention mechanism in a feature fusion network, and using an SIoU position loss function by a detection head; a YOLOv5s network model is trained and tested; and inputting a picture of the unlabeled detection target to obtain a detection target result. According to the method, the GAM attention mechanism is added in the feature fusion network, so that the model is more focused on features beneficial to target detection, and the detection accuracy is improved. Secondly, a position loss function is replaced by SIoU, so that the convergence speed of the model and the reasoning accuracy are improved; the improved YOLOv5s is used for target detection, so that the omission ratio and the false detection rate of target detection are reduced to a certain extent.

The invention belongs to the technical field of unmanned aerial vehicle attack and defense, and discloses an intelligent defense and attack control system for an unmanned aerial vehicle. The method comprises the following steps: acquiring environment sensing data of an unmanned aerial vehicle, wherein the environment sensing data comprises state data of the unmanned aerial vehicle and image data of a surrounding environment; preprocessing the image data of the surrounding environment to obtain model input image data; performing target detection on the model input image data by using a variant Poolform algorithm to obtain comprehensive information of a target object; according to the state data of the unmanned aerial vehicle and the comprehensive information of the target object, solving an optimal attack defense strategy by using a perception integrated attack algorithm; the optimal attack defense strategy is sent to an unmanned aerial vehicle execution terminal, and the unmanned aerial vehicle is controlled to execute the optimal attack defense strategy; the optimal attack and defense strategy can be dynamically generated according to the real-time change of the environment situation, and the survivability and the attack efficiency of the unmanned aerial vehicle in the confrontation environment are remarkably improved.

The invention discloses a bridge crack detection method based on image recognition, and belongs to the technical field of image recognition, and the method comprises the steps: firstly, dividing a bridge image into three single-color channels of red, green and blue, carrying out the clustering and scanning of each spectral image, and precisely positioning a suspected crack point; a red channel weight matrix, a green channel weight matrix and a blue channel weight matrix are constructed by calculating a channel contrast value and a region track distortion value, and then gray processing and brightness adjustment are performed on a bridge image to enhance a crack region. Then, texture enhancement is carried out on the adjusted images through the red channel weight matrix, the green channel weight matrix and the blue channel weight matrix, and three texture-enhanced images are generated; and finally, processing the enhanced image by using a three-channel crack identification neural network to obtain the bridge crack, thereby improving the accuracy of crack detection.

The invention relates to a traffic small target detection method and device for a high-speed driving vehicle, equipment and a medium, and relates to the field of image processing, and the method comprises the steps: inputting a traffic scene image frame into a second traffic small target detection model which is trained to a convergence state, and carrying out the detection of a traffic scene image frame in a plurality of multi-layer stacked recursive Star Block modules of a StarNet lightweight network, performing element-by-element multiplication on image features corresponding to the traffic scene image frames by adopting star operation to exponentially increase implicit feature dimensions in the traffic scene image frames so as to determine high-dimensional image features of the traffic scene image frames; and inputting the high-dimensional image features into a detection head network in a traffic small target detection model to obtain a plurality of traffic small targets in the traffic scene image frame so as to complete traffic small target detection of the high-speed driving vehicle. According to the method, the small traffic target can be accurately and quickly detected in a traffic scene in high-speed driving, and higher target detection performance can be realized in a resource-limited environment such as high-speed driving vehicles.

The invention relates to the field of fault detection for display screen production, in particular to a fault detection system for display screen production, which comprises a personnel login module, a data acquisition module, an electrical parameter analysis and detection module, a signal analysis module, a control processing module and a remote monitoring module, the data acquisition module determines fault characteristics of a signal source, the electrical parameter analysis and detection module receives data information sent by the data acquisition module and analyzes electrical parameter faults of the LED display screen, and the remote monitoring module monitors damage to the outer surface of the display screen according to a high-definition camera and removes the LED display screen with damage flaws. The invention aims to realize real-time, comprehensive and accurate testing of electrical parameters, signals and appearance flaws, timely discover potential fault risks and improve production quality and efficiency.

The invention discloses a coal mine shaft crack identification method in a non-illumination environment based on an optical fiber sensing grating, and belongs to the field of coal mine shaft disease detection. According to the method, distributed optical fibers form an optical fiber sensing grating, the optical fiber sensing grating is fixed on the surface of a well wall, an optical fiber demodulator is used for collecting strain along the optical fibers, a blank blind area of the optical fiber sensing grating is filled through an interpolation algorithm, and a strain field image data set of the well wall is obtained; and a CCD industrial camera is matched to shoot a well wall crack image sample at the optical fiber sensing grating, binarization processing is carried out, and a well wall crack image data set is established. And establishing a well wall crack identification model based on deep learning, introducing two parallel feature extraction branches, respectively receiving an optical fiber strain field image data set and a well wall crack image data set, and performing feature extraction and fusion of the two different types of data sets through training. Because the strain data is sensed through the optical fiber, the illumination environment is not needed, and after the model training is completed, the identification and early warning of the well wall crack in the illumination-free environment can be realized only by inputting the optical fiber strain data acquired in real time into the trained model.

The invention discloses an optical remote sensing image saliency target detection model and detection method, and belongs to the technical field of target detection. The saliency target detection model comprises a PVT backbone network module, an edge information extraction and fusion module BG, a feature enhancement module HAM, two space attention modules, a feature interaction module KTM and a feature fusion module. According to the invention, by introducing an edge detection module (BG) and a feature enhancement module (HAM), the target detection precision under a complex background is effectively improved. Compared with the existing CNN-based method, the method provided by the invention can better process targets with different scales and direction changes, reduces the probability of missing detection and false detection, and enables saliency target detection to be more complete and accurate.

The invention discloses a detection and tracking method and device for dense pedestrians, relates to the technical field of computer vision, and can solve the problem of low tracking accuracy during real-time detection and tracking in a scene with extremely dense pedestrians in the prior art. According to the specific technical scheme, firstly, a dense pedestrian detection data set and a tracking data set are obtained; an alpha-EIoU loss, a GAM attention mechanism and a Swin Transform module are used for improving a YOLOv5s algorithm, and a detection model is obtained; training and optimizing the detection model to obtain a target detection model; a ShuffleNetV2 module is introduced to improve a DeepSORT algorithm, and then training is performed to obtain a target tracking model; and detecting and tracking the pedestrian target video according to the target detection model and the target tracking model to obtain an experimental result of detection and tracking. The method is used for detecting and tracking dense pedestrians.

The invention discloses a tunnel fire detection and simulation method and system driven by video data, belongs to the technical field of computer vision and machine learning, and solves the problem that the current tunnel fire target detection algorithm is easy to cause insufficient distinguishing capability of vehicles and flames in a complex multi-target scene environment. The method comprises the following steps: preprocessing an obtained original data set to obtain a target data set to train an improved YOLOv8 neural network model, and carrying out target detection on a to-be-detected tunnel fire scene disaster image to obtain a central coordinate of each preliminary disaster object target for correction; after correction, a monocular vision positioning method is adopted to carry out position calibration and distinguishing on the corrected disaster situation object target, and then real-time scene twinborn modeling is carried out, so that a tunnel fire scene twinborn model is obtained and dynamically updated; and performing flame spread analysis on the tunnel fire scene based on the tunnel fire scene twinborn model, and realizing dynamic updating in un ity. The method is used for tunnel fire detection and simulation driven by video data.