Research On Image Anomaly Detection And Energy Consumption Optimization Based On Internet Of Things

At present,it has entered the rapid development stage of the Internet of Things.The Internet of Things is widely used in combination with intelligent transportation for video traffic monitoring.The negative impact of traffic safety hazards in current life is becoming more serious.In the scenarios where vehicles are prohibited from driving in pedestrian streets,campuses and other scenes,manual monitoring is prone to "missing reports" and "false reports",which may lead to chaos and harm people's life safety,so the automatic detection of abnormal vehicles has practical significance.At present,most Io T nodes still use nonrechargeable power sources for power supply,which causes network energy consumption to directly affect the length of the network life cycle.Due to network energy exhaustion,monitoring is interrupted,and pedestrians cannot be provided with timely reminders,endangering people's lives Therefore,more and more researchers are doing a lot of research on energy consumption reduction of the Internet of Things.Firstly,aiming at the problems of ghosting and cavitation in foreground object detection using Gaussian mixture model method,an abnormal vehicle detection algorithm based on SSIM(structural similarity)to improve Gaussian mixture model is proposed.Exponential function is introduced to optimize the weight updating process in Gaussian modeling process,which improves the updating speed.SSIM is used to calculate the similarity between the pixel points of the two images,and the improved Gaussian mixture model is used to reset the pixel values in combination with SSIM to obtain the foreground image.Using the graphic handle function to optimize the connected domain method to detect abnormal vehicles in the foreground region,it is possible to detect abnormal vehicles more accurately while marking frames closer to the shape of the vehicle.Secondly,aiming at the problem that the grey wolf optimization algorithm is easy to converge to the local optimal value,an improved grey wolf optimization algorithm is proposed.By introducing an exponential function into the convergence factor,the grey wolf optimization process is optimized so that it has a nonlinear approximation step size and jumps out of the local optimal solution.By adopting a dynamic weight position updating strategy based on Cauchy mutation,the algorithm is further prevented from falling into a local optimal value problem,and the solution speed of the algorithm is improved.Thirdly,aiming at the problem that the energy consumption of the Internet of Things network is too large because the compressed image data is still huge when the image is compressed and transmitted to the cloud,the abnormal image is converted into instruction information and then transmitted to the cloud to reduce the transmission data amount.Reprocessing the result of anomaly detection at the local terminal,adding the node ID,time,judgment instruction,check information and other information to generate message data containing the anomaly detection result,uploading the message data with small data amount to the cloud through the Internet of Things,and the cloud can carry out anomaly judgment and warn abnormal vehicles.The method can effectively reduce the transmission energy consumption of the Internet of Things network.Finally,aiming at the problem of excessive network energy consumption caused by randomly selecting cluster head nodes in traditional clustering algorithms,an improved LEACH algorithm based on improved gray wolf optimization algorithm is proposed.The improved gray wolf optimization algorithm is adopted to optimize the cluster head node selection process,and the comprehensive factors such as cluster head node residual energy,network energy,node location information and node link connectivity are taken as judgment indexes for cluster head node selection,so as to achieve the aims of reducing the energy consumption of the Internet of Things network and prolonging the life cycle of the Internet of Things network.