Aiming at the problems that the classical Gaussian mixture model is unable to detect the complete moving object, and is sensitive to the light mutation scenes and so on, an improved algorithm is proposed for moving ob...Aiming at the problems that the classical Gaussian mixture model is unable to detect the complete moving object, and is sensitive to the light mutation scenes and so on, an improved algorithm is proposed for moving object detection based on Gaussian mixture model and three-frame difference method. In the process of extracting the moving region, the improved three-frame difference method uses the dynamic segmentation threshold and edge detection technology, and it is first used to solve the problems such as the illumination mutation and the discontinuity of the target edge. Then, a new adaptive selection strategy of the number of Gaussian distributions is introduced to reduce the processing time and improve accuracy of detection. Finally, HSV color space is used to remove shadow regions, and the whole moving object is detected. Experimental results show that the proposed algorithm can detect moving objects in various situations effectively.展开更多
Moving object detection including background subtraction and morphological processing is a critical research topic for video surveillance because of its high computational loading and power consumption. This paper pro...Moving object detection including background subtraction and morphological processing is a critical research topic for video surveillance because of its high computational loading and power consumption. This paper proposes a hardware design to accelerate the computation of background subtraction with low power consumption. A real-time background subtraction method is designed with a frame-buffer scheme and function partition to improve throughput, and implemented using Verilog HDL on FPGA. The design parallelizes the computations of background update and subtraction with a seven-stage pipeline. A stripe-based morphological processing and accounting for the completion of detected objects is devised. Simulation results for videos of VGA resolutions on a low-end FPGA device show 368 fps throughput for only the real-time background subtraction module, and 51 fps for the whole system, including off-chip memory access. Real-time efficiency with low power consumption and low resource utilization is thus demonstrated.展开更多
Simultaneous localisation and mapping(SLAM)are the basis for many robotic applications.As the front end of SLAM,visual odometry is mainly used to estimate camera pose.In dynamic scenes,classical methods are deteriorat...Simultaneous localisation and mapping(SLAM)are the basis for many robotic applications.As the front end of SLAM,visual odometry is mainly used to estimate camera pose.In dynamic scenes,classical methods are deteriorated by dynamic objects and cannot achieve satisfactory results.In order to improve the robustness of visual odometry in dynamic scenes,this paper proposed a dynamic region detection method based on RGBD images.Firstly,all feature points on the RGB image are classified as dynamic and static using a triangle constraint and the epipolar geometric constraint successively.Meanwhile,the depth image is clustered using the K-Means method.The classified feature points are mapped to the clustered depth image,and a dynamic or static label is assigned to each cluster according to the number of dynamic feature points.Subsequently,a dynamic region mask for the RGB image is generated based on the dynamic clusters in the depth image,and the feature points covered by the mask are all removed.The remaining static feature points are applied to estimate the camera pose.Finally,some experimental results are provided to demonstrate the feasibility and performance.展开更多
Purpose-The purpose of this paper is to meet the large demand for the new-generation intelligence monitoring systems that are used to detect targets within a dynamic background.Design/methodology/approach-A dynamic ta...Purpose-The purpose of this paper is to meet the large demand for the new-generation intelligence monitoring systems that are used to detect targets within a dynamic background.Design/methodology/approach-A dynamic target detection method based on the fusion of optical flow and neural network is proposed.Findings-Simulation results verify the accuracy of the moving object detection based on optical flow andneural network fusion.Themethod eliminates the influence caused bythe movement of thecamera to detect the target and has the ability to extract a complete moving target.Practical implications-It provides a powerful safeguard for target detection and targets the tracking application.Originality/value-The proposed method represents the fusion of optical flow and neural network to detect the moving object,and it can be used in new-generation intelligent monitoring systems.展开更多
This study focuses on developing deep learning methods for small and dim target detection.We model infrared images as the union of the target region and background region.Based on this model,the target detection probl...This study focuses on developing deep learning methods for small and dim target detection.We model infrared images as the union of the target region and background region.Based on this model,the target detection problem is considered a two‐class segmentation problem that divides an image into the target and background.Therefore,a neural network called SDDNet for single‐frame images is constructed.The network yields target extraction results according to the original images.For multiframe images,a network called IC‐SDDNet,a combination of SDDNet and an interframe correlation network module is constructed.SDDNet and IC‐SDDNet achieve target detection rates close to 1 on typical datasets with very low false positives,thereby performing significantly better than current methods.Both models can be executed end to end,so both are very convenient to use,and their implementation efficiency is very high.Average speeds of 540+/230+FPS and 170+/60+FPS are achieved with SDDNet and IC‐SDDNet on a single Tesla V100 graphics processing unit and a single Jetson TX2 embedded module respectively.Additionally,neither network needs to use future information,so both networks can be directly used in real‐time systems.The well‐trained models and codes used in this study are available at https://github.com/LittlePieces/ObjectDetection.展开更多
There are an increasing number of Narrow Band IoT devices being manufactured as the technology behind them develops quickly.The high co‐channel interference and signal attenuation seen in edge Narrow Band IoT devices...There are an increasing number of Narrow Band IoT devices being manufactured as the technology behind them develops quickly.The high co‐channel interference and signal attenuation seen in edge Narrow Band IoT devices make it challenging to guarantee the service quality of these devices.To maximise the data rate fairness of Narrow Band IoT devices,a multi‐dimensional indoor localisation model is devised,consisting of transmission power,data scheduling,and time slot scheduling,based on a network model that employs non‐orthogonal multiple access via a relay.Based on this network model,the optimisation goal of Narrow Band IoT device data rate ratio fairness is first established by the authors,while taking into account the Narrow Band IoT network:The multidimensional indoor localisation optimisation model of equipment tends to minimize data rate,energy constraints and EH relay energy and data buffer constraints,data scheduling and time slot scheduling.As a result,each Narrow Band IoT device's data rate needs are met while the network's overall performance is optimised.We investigate the model's potential for convex optimisation and offer an algorithm for optimising the distribution of multiple resources using the KKT criterion.The current work primarily considers the NOMA Narrow Band IoT network under a single EH relay.However,the growth of Narrow Band IoT devices also leads to a rise in co‐channel interference,which impacts NOMA's performance enhancement.Through simulation,the proposed approach is successfully shown.These improvements have boosted the network's energy efficiency by 44.1%,data rate proportional fairness by 11.9%,and spectrum efficiency by 55.4%.展开更多
With the explosive growth of surveillance video data,browsing videos quickly and effectively has become an urgent problem.Video key frame extraction has received widespread attention as an effective solution.However,a...With the explosive growth of surveillance video data,browsing videos quickly and effectively has become an urgent problem.Video key frame extraction has received widespread attention as an effective solution.However,accurately capturing the local motion state changes of moving objects in the video is still challenging in key frame extraction.The target center offset can reflect the change of its motion state.This observation proposed a novel key frame extraction method based on moving objects center offset in this paper.The proposed method utilizes the center offset to obtain the global and local motion state information of moving objects,and meanwhile,selects the video frame where the center offset curve changes suddenly as the key frame.Such processing effectively overcomes the inaccuracy of traditional key frame extraction methods.Initially,extracting the center point of each frame.Subsequently,calculating the center point offset of each frame and forming the center offset curve by connecting the center offset of each frame.Finally,extracting candidate key frames and optimizing them to generate final key frames.The experimental results demonstrate that the proposed method outperforms contrast methods to capturing the local motion state changes of moving objects.展开更多
Aiming at the fixed-view video surveillance scene,this paper proposes a video object detection method that combines motion features and YOLO.The method uses the method of filtering video frames without motion features...Aiming at the fixed-view video surveillance scene,this paper proposes a video object detection method that combines motion features and YOLO.The method uses the method of filtering video frames without motion features and segmenting video frames with motion features to reduce the reasoning pressure of the YOLO algorithm model.In this process,video frames containing moving objects are first obtained by the moving object detection module.Second,the moving target will be recognized by the object of interest recognition module.Finally,the background decision module records and analyzes the detection results to obtain background model updates or result output.It detects moving objects without using traditional background modeling methods.Experiments based on theCDnet2014 dataset showthat our method improves the missed detection rate by 0.098% and the average inference speed per frame by 45.62%compared with the YOLO-based humanoid detection method.Furthermore,the method has superior performance in scenarios where target objects appear less frequently(substations,transmission lines,and hazardous areas).展开更多
Following the popularity of digital video application,video copying and dissemination have become very easy;however,this makes video hacking and piracy a potential threat in video communication.Video watermarking tech...Following the popularity of digital video application,video copying and dissemination have become very easy;however,this makes video hacking and piracy a potential threat in video communication.Video watermarking technology can solve the problem of copyright protection,and thus,it has been extensively researched.The robustness of the video watermarking algorithm in the bitstream domain is poor,especially the anti-recompression ability,since the watermarked video may be compressed again before transmitting.Considering this,this paper proposes a video watermarking algorithm in the bitstream domain based on moving object detection.To increase the robustness of the watermarking scheme,the moving macroblocks that belong to the moving object in each P frame can be identified via the moving object detection algorithm.Then,watermark embedding in the moving macroblocks is performed using codeword substitution to ensure the consistency of the bitstream.Moving object detection and watermark embedding are independent and are both carried out in the bitstream domain by partially decoding the bitstream;this avoids the complete decoding and reconstruction of the video,making the method to be highly efficient.The simulation results confirm that the proposed method is robust against recompression and has little impact on the video visual quality and no influence on the bit rate.展开更多
文摘Aiming at the problems that the classical Gaussian mixture model is unable to detect the complete moving object, and is sensitive to the light mutation scenes and so on, an improved algorithm is proposed for moving object detection based on Gaussian mixture model and three-frame difference method. In the process of extracting the moving region, the improved three-frame difference method uses the dynamic segmentation threshold and edge detection technology, and it is first used to solve the problems such as the illumination mutation and the discontinuity of the target edge. Then, a new adaptive selection strategy of the number of Gaussian distributions is introduced to reduce the processing time and improve accuracy of detection. Finally, HSV color space is used to remove shadow regions, and the whole moving object is detected. Experimental results show that the proposed algorithm can detect moving objects in various situations effectively.
文摘Moving object detection including background subtraction and morphological processing is a critical research topic for video surveillance because of its high computational loading and power consumption. This paper proposes a hardware design to accelerate the computation of background subtraction with low power consumption. A real-time background subtraction method is designed with a frame-buffer scheme and function partition to improve throughput, and implemented using Verilog HDL on FPGA. The design parallelizes the computations of background update and subtraction with a seven-stage pipeline. A stripe-based morphological processing and accounting for the completion of detected objects is devised. Simulation results for videos of VGA resolutions on a low-end FPGA device show 368 fps throughput for only the real-time background subtraction module, and 51 fps for the whole system, including off-chip memory access. Real-time efficiency with low power consumption and low resource utilization is thus demonstrated.
基金supported in part by the National Natural Science Foundation of China(Grant No.U1913201,U22B2041)Natural Science Foundation of Liaoning Province(Grant No.2019-ZD-0169).
文摘Simultaneous localisation and mapping(SLAM)are the basis for many robotic applications.As the front end of SLAM,visual odometry is mainly used to estimate camera pose.In dynamic scenes,classical methods are deteriorated by dynamic objects and cannot achieve satisfactory results.In order to improve the robustness of visual odometry in dynamic scenes,this paper proposed a dynamic region detection method based on RGBD images.Firstly,all feature points on the RGB image are classified as dynamic and static using a triangle constraint and the epipolar geometric constraint successively.Meanwhile,the depth image is clustered using the K-Means method.The classified feature points are mapped to the clustered depth image,and a dynamic or static label is assigned to each cluster according to the number of dynamic feature points.Subsequently,a dynamic region mask for the RGB image is generated based on the dynamic clusters in the depth image,and the feature points covered by the mask are all removed.The remaining static feature points are applied to estimate the camera pose.Finally,some experimental results are provided to demonstrate the feasibility and performance.
基金This work was supported by the National Natural Science Foundation of China(No.61304223,No.61673209 and No.61533008)the Fundamental Research Funds for the Central Universities(No.NZ2015206 and No.NJ20160026).
文摘Purpose-The purpose of this paper is to meet the large demand for the new-generation intelligence monitoring systems that are used to detect targets within a dynamic background.Design/methodology/approach-A dynamic target detection method based on the fusion of optical flow and neural network is proposed.Findings-Simulation results verify the accuracy of the moving object detection based on optical flow andneural network fusion.Themethod eliminates the influence caused bythe movement of thecamera to detect the target and has the ability to extract a complete moving target.Practical implications-It provides a powerful safeguard for target detection and targets the tracking application.Originality/value-The proposed method represents the fusion of optical flow and neural network to detect the moving object,and it can be used in new-generation intelligent monitoring systems.
文摘This study focuses on developing deep learning methods for small and dim target detection.We model infrared images as the union of the target region and background region.Based on this model,the target detection problem is considered a two‐class segmentation problem that divides an image into the target and background.Therefore,a neural network called SDDNet for single‐frame images is constructed.The network yields target extraction results according to the original images.For multiframe images,a network called IC‐SDDNet,a combination of SDDNet and an interframe correlation network module is constructed.SDDNet and IC‐SDDNet achieve target detection rates close to 1 on typical datasets with very low false positives,thereby performing significantly better than current methods.Both models can be executed end to end,so both are very convenient to use,and their implementation efficiency is very high.Average speeds of 540+/230+FPS and 170+/60+FPS are achieved with SDDNet and IC‐SDDNet on a single Tesla V100 graphics processing unit and a single Jetson TX2 embedded module respectively.Additionally,neither network needs to use future information,so both networks can be directly used in real‐time systems.The well‐trained models and codes used in this study are available at https://github.com/LittlePieces/ObjectDetection.
文摘There are an increasing number of Narrow Band IoT devices being manufactured as the technology behind them develops quickly.The high co‐channel interference and signal attenuation seen in edge Narrow Band IoT devices make it challenging to guarantee the service quality of these devices.To maximise the data rate fairness of Narrow Band IoT devices,a multi‐dimensional indoor localisation model is devised,consisting of transmission power,data scheduling,and time slot scheduling,based on a network model that employs non‐orthogonal multiple access via a relay.Based on this network model,the optimisation goal of Narrow Band IoT device data rate ratio fairness is first established by the authors,while taking into account the Narrow Band IoT network:The multidimensional indoor localisation optimisation model of equipment tends to minimize data rate,energy constraints and EH relay energy and data buffer constraints,data scheduling and time slot scheduling.As a result,each Narrow Band IoT device's data rate needs are met while the network's overall performance is optimised.We investigate the model's potential for convex optimisation and offer an algorithm for optimising the distribution of multiple resources using the KKT criterion.The current work primarily considers the NOMA Narrow Band IoT network under a single EH relay.However,the growth of Narrow Band IoT devices also leads to a rise in co‐channel interference,which impacts NOMA's performance enhancement.Through simulation,the proposed approach is successfully shown.These improvements have boosted the network's energy efficiency by 44.1%,data rate proportional fairness by 11.9%,and spectrum efficiency by 55.4%.
基金This work was supported by the National Nature Science Foundation of China(Grant No.61702347,61772225)Natural Science Foundation of Hebei Province(Grant No.F2017210161).
文摘With the explosive growth of surveillance video data,browsing videos quickly and effectively has become an urgent problem.Video key frame extraction has received widespread attention as an effective solution.However,accurately capturing the local motion state changes of moving objects in the video is still challenging in key frame extraction.The target center offset can reflect the change of its motion state.This observation proposed a novel key frame extraction method based on moving objects center offset in this paper.The proposed method utilizes the center offset to obtain the global and local motion state information of moving objects,and meanwhile,selects the video frame where the center offset curve changes suddenly as the key frame.Such processing effectively overcomes the inaccuracy of traditional key frame extraction methods.Initially,extracting the center point of each frame.Subsequently,calculating the center point offset of each frame and forming the center offset curve by connecting the center offset of each frame.Finally,extracting candidate key frames and optimizing them to generate final key frames.The experimental results demonstrate that the proposed method outperforms contrast methods to capturing the local motion state changes of moving objects.
文摘Aiming at the fixed-view video surveillance scene,this paper proposes a video object detection method that combines motion features and YOLO.The method uses the method of filtering video frames without motion features and segmenting video frames with motion features to reduce the reasoning pressure of the YOLO algorithm model.In this process,video frames containing moving objects are first obtained by the moving object detection module.Second,the moving target will be recognized by the object of interest recognition module.Finally,the background decision module records and analyzes the detection results to obtain background model updates or result output.It detects moving objects without using traditional background modeling methods.Experiments based on theCDnet2014 dataset showthat our method improves the missed detection rate by 0.098% and the average inference speed per frame by 45.62%compared with the YOLO-based humanoid detection method.Furthermore,the method has superior performance in scenarios where target objects appear less frequently(substations,transmission lines,and hazardous areas).
基金supported by the National Natural Science Foundation of China(No.61702563)。
文摘Following the popularity of digital video application,video copying and dissemination have become very easy;however,this makes video hacking and piracy a potential threat in video communication.Video watermarking technology can solve the problem of copyright protection,and thus,it has been extensively researched.The robustness of the video watermarking algorithm in the bitstream domain is poor,especially the anti-recompression ability,since the watermarked video may be compressed again before transmitting.Considering this,this paper proposes a video watermarking algorithm in the bitstream domain based on moving object detection.To increase the robustness of the watermarking scheme,the moving macroblocks that belong to the moving object in each P frame can be identified via the moving object detection algorithm.Then,watermark embedding in the moving macroblocks is performed using codeword substitution to ensure the consistency of the bitstream.Moving object detection and watermark embedding are independent and are both carried out in the bitstream domain by partially decoding the bitstream;this avoids the complete decoding and reconstruction of the video,making the method to be highly efficient.The simulation results confirm that the proposed method is robust against recompression and has little impact on the video visual quality and no influence on the bit rate.