Segmentation Based Real Time Anomaly Detection and Tracking Model for Pedestrian Walkways

Presently,video surveillance is commonly employed to ensure security in public places such as traffic signals,malls,railway stations,etc.A major chal-lenge in video surveillance is the identification of anomalies that exist in it such as crimes,thefts,and so on.Besides,the anomaly detection in pedestrian walkways has gained significant attention among the computer vision communities to enhance pedestrian safety.The recent advances of Deep Learning(DL)models have received considerable attention in different processes such as object detec-tion,image classification,etc.In this aspect,this article designs a new Panoptic Feature Pyramid Network based Anomaly Detection and Tracking(PFPN-ADT)model for pedestrian walkways.The proposed model majorly aims to the recognition and classification of different anomalies present in the pedestrian walkway like vehicles,skaters,etc.The proposed model involves panoptic seg-mentation model,called Panoptic Feature Pyramid Network(PFPN)is employed for the object recognition process.For object classification,Compact Bat Algo-rithm(CBA)with Stacked Auto Encoder(SAE)is applied for the classification of recognized objects.For ensuring the enhanced results better anomaly detection performance of the PFPN-ADT technique,a comparison study is made using Uni-versity of California San Diego(UCSD)Anomaly data and other benchmark data-sets(such as Cityscapes,ADE20K,COCO),and the outcomes are compared with the Mask Recurrent Convolutional Neural Network(RCNN)and Faster Convolu-tional Neural Network(CNN)models.The simulation outcome demonstrated the enhanced performance of the PFPN-ADT technique over the other methods.

An extended social force model on unidirectional flow considering psychological and behavioral impacts of hazard source

An accurate assessment of the evacuation efficiency in case of disasters is of vital importance to the safety design of buildings and street blocks.Hazard sources not only physically but psychologically affect the pedestrians,which may further alter their behavioral patterns.This effect is especially significant in narrow spaces,such as corridors and alleys.This study aims to integrate a non-spreading hazard source into the social force model following the results from a previous experiment and simulation,and to simulate unidirectional pedestrian flows over various crowd densities and clarity–intensity properties of the hazard source.The integration include a virtual repulsion force from the hazard source and a decay on the social force term.The simulations reveal(i)that the hazard source creates virtual bottlenecks that suppress the flow,(ii)that the inter-pedestrian push forms a stabilisation phase on the flow-density curve within medium-to-high densities,and(iii)that the pedestrians are prone to a less orderly and stable pattern of movement in low clarity–intensity scenarios,possibly with lateral collisions passing the hazard source.

Hybrid pedestrian positioning system using wearable inertial sensors and ultrasonic ranging

Pedestrian positioning system(PPS)using wearable inertial sensors has wide applications towards various emerging fields such as smart healthcare,emergency rescue,soldier positioning,etc.The performance of traditional PPS is limited by the cumulative error of inertial sensors,complex motion modes of pedestrians,and the low robustness of the multi-sensor collaboration structure.This paper presents a hybrid pedestrian positioning system using the combination of wearable inertial sensors and ultrasonic ranging(H-PPS).A robust two nodes integration structure is developed to adaptively combine the motion data acquired from the single waist-mounted and foot-mounted node,and enhanced by a novel ellipsoid constraint model.In addition,a deep-learning-based walking speed estimator is proposed by considering all the motion features provided by different nodes,which effectively reduces the cumulative error originating from inertial sensors.Finally,a comprehensive data and model dual-driven model is presented to effectively combine the motion data provided by different sensor nodes and walking speed estimator,and multi-level constraints are extracted to further improve the performance of the overall system.Experimental results indicate that the proposed H-PPS significantly improves the performance of the single PPS and outperforms existing algorithms in accuracy index under complex indoor scenarios.

Modeling walking behavior of pedestrian groups with floor field cellular automaton approach

Walking in groups is very common in a realistic walking environment. An extended floor field cellular automaton （CA） model is therefore proposed to describe the walking behavior of pedestrian groups, This model represents the motion of pedestrian groups in a realistic way. The simulation results reveal that the walking behavior of groups has an important but negative influence on pedestrian flow dynamics, especially when the density is at a high level. The presence of pedestrian groups retards the emergence of lane formation and increases the instability of operation of pedestrian flow. Moreover, the average velocity and volume of pedestrian flow are significantly reduced due to the group motion. Meanwhile, the parameter-sensitive analysis suggests that pedestrian groups should make a compromise between efficient movement and staying coherent with a certain spatial structure when walking in a dense crowd.

A new collision avoidance model for pedestrian dynamics

The pedestrians can only avoid collisions passively under the action of forces during simulations using the social force model, which may lead to unnatural behaviors. This paper proposes an optimization-based model for the avoidance of collisions, where the social repulsive force is removed in favor of a search for the quickest path to destination in the pedestrian＇s vision field. In this way, the behaviors of pedestrians are governed by changing their desired walking direction and desired speed. By combining the critical factors of pedestrian movement, such as positions of the exit and obstacles and velocities of the neighbors, the choice of desired velocity has been rendered to a discrete optimization problem. Therefore,it is the self-driven force that leads pedestrians to a free path rather than the repulsive force, which means the pedestrians can actively avoid collisions. The new model is verified by comparing with the fundamental diagram and actual data. The simulation results of individual avoidance trajectories and crowd avoidance behaviors demonstrate the reasonability of the proposed model.

Cellular automata modeling of pedestrian＇s crossing dynamics

Cellular automata modeling techniques and the characteristics of mixed traffic flow were used to derive the 2-dimensional model presented here for simulation of pedestrian’s crossing dynamics. A conception of “stop point” is introduced to deal with traffic obstacles and resolve conflicts among pedestrians or between pedestrians and the other vehicles on the crosswalk. The model can be easily extended, is very efficient for simulation of pedestrian’s crossing dy- namics, can be integrated into traffic simulation software, and has been proved feasible by simulation experiments.

Microscopic dynamic simulation model for pedestrian at signalized intersection

Pedestrian's road-crossing model is the key part of micro-simulation for mixed traffic at signalized intersection.To reproduce the crossing behavior of pedestrians,the microscopic behaviors of the pedestrians passing through the crosswalk at signalized intersection were analyzed.A pedestrian's decision making model based on gap acceptance theory was proposed.Based on the field data at three typical intersections in Beijing,China,the critical gaps and lags of pedestrians were calibrated.In addition,considering pedestrian's required space,a modification of the social force model that consists of a self-deceleration mechanism prevents a simulated pedestrian from continuously pushing over other pedestrians,making the simulation more realistic.After the simple change,the modified social force model is able to reproduce the fundamental diagram of pedestrian flows for densities less than 3.5 m-2 as reported in the literature.

The Korteweg-de Vires equation for the bidirectional pedestrian flow model considering the next-nearest-neighbor effect

This paper focuses on a two-dimensional bidirectional pedestrian flow model which involves the next-nearest-neighbor effect. The stability condition and the Korteweg-de Vries （KdV） equation are derived to describe the density wave of pedestrian congestion by linear stability and nonlinear analysis. Through theoretical analysis, the soliton solution is obtained.

Cyber-physical Modeling and Control of Crowd of Pedestrians:A Review and New Framework

Recent advances in modeling and control of crowd of pedestrians are briefly surveyed in this paper. Possibilities of applying fractional calculus in the modeling of crowd of pedestrians have been shortly reviewed and discussed from different aspects such as descriptions of motion, interactions of long range and effects of memory. Control of the crowd of pedestrians have also been formulated using the framework of cyber-physical systems and been realized using networked Segways with onboard emergency response personnels to regulate the velocity and flux of the crowd. Platform for verification of the theoretical results are also provided in this paper. © 2014 Chinese Association of Automation.

Simulation based on a modified social force model for sensitivity to emergency signs in subway station

The subway is the primary travel tool for urban residents in China. Due to the complex structure of the subway and high personnel density in rush hours, subway evacuation capacity is critical. The subway evacuation model is explored in this work by combining the improved social force model with the view radius using the Vicsek model. The pedestrians are divided into two categories based on different force models. The first category is sensitive pedestrians who have normal responses to emergency signs. The second category is insensitive pedestrians. By simulating different proportions of the insensitive pedestrians, we find that the escape time is directly proportional to the number of insensitive pedestrians and inversely proportional to the view radius. However, when the view radius is large enough, the escape time does not change significantly, and the evacuation of people in a small view radius environment tends to be integrated. With the improvement of view radius conditions, the escape time changes more obviously with the proportion of insensitive pedestrians. A new emergency sign layout is proposed, and the simulations show that the proposed layout can effectively reduce the escape time in a small view radius environment. However, the evacuation effect of the new escape sign layout on the large view radius environment is not apparent. In this case, the exit setting emerges as an additional factor affecting the escape time.

Time-dependent Ginzburg-Landau equation for lattice hydrodynamic model describing pedestrian flow

A thermodynamic theory is formulated to describe the phase transition and critical phenomena in pedestrian flow. Based on the extended lattice hydrodynamic pedestrian model taking the interaction of the next-nearest-neighbor persons into account, the time-dependent Ginzburg-Landau （TDGL） equation is derived to describe the pedestrian flow near the critical point through the nonlinear analysis method. The corresponding two solutions, the uniform and the kink solutions, are given. The coexisting curve, spinodal line, and critical point are obtained by the first and second derivatives of the thermodynamic potential.

A modified heuristics-based model for simulating realistic pedestrian movement behavior

Pedestrian movement simulation models are used in various areas,such as building evacuation,transportation engi-neering,and safety management of large events.It also provides effective means to uncover underlying mechanisms of collective behaviors.In this work,a modified heuristics-based model is presented.In this model,the potential collisions in the moving process are explicitly considered.Meanwhile,a series of simulations is conducted in two typical scenarios to demonstrate the influence of critical parameters on model performance.It is found that when facing a wide obstacle in a corridor,the larger the visual radius,the earlier the pedestrian starts to make a detour.In addition,when a pedestrian observes a large crowd walking toward him,he chooses to make a detour and moves in the flow in a uniform direction.Furthermore,the model can reproduce the lane formation pedestrian flow phenomena in relatively high-density situations.With the increase of pedestrian visual radius and the weight of potential collision resistance,more stable pedestrian lanes and fewer moving-through-the-counterflow pedestrians can be observed.In terms of model validation,the density-speed relationship of simulation results accords well with that of the published empirical data.Our results demonstrate that the modified heuristics-based model can overcome the deficiency of the original model,and reproduce more realistic pedestrian movement behavior.

Simulation of High Density Pedestrian Flow: A Microscopic Model

In recent years, modelling crowd and evacuation dynamics has become very important, with increasing huge numbers of people gathering around the world for many reasons and events. The fact that our global population grows dramatically every year and the current public transport systems are able to transport large amounts of people heightens the risk of crowd panic or crush. Pedestrian models are based on macroscopic or microscopic behaviour. In this paper, we are interested in developing models that can be used for evacuation control strategies. This model will be based on microscopic pedestrian simulation models, and its evolution and design requires a lot of information and data. The people stream will be simulated, based on mathematical models derived from empirical data about pedestrian flows. This model is developed from image data bases, so called empirical data, taken from a video camera or data obtained using human detectors. We consider the individuals as autonomous particles interacting through social and physical forces, which is an approach that has been used to simulate crowd behaviour. The target of this work is to describe a comprehensive approach to model a huge number of pedestrians and to simulate high density crowd behaviour in overcrowding places, e.g. sport, concert and pilgrimage places, and to assist engineering in the resolution of complicated problems through integrating a number of models from different research domains.

Dynamic Modeling of Pedestrian Legform and Prediction of Pedestrian Injuries

Decreasing the death toll of pedestrians in traffic accidents is one of the most urgent tasks to be solved all over the world. This paper describes the prediction of pedestrian injuries for the TRL legform impactor using MATLAB. The TRL legform impactor consists of three parts: a femur, a tibia, and a ligament connecting them. The impactor was physically modelled with springs, dampers and two masses as a dynamic model. The impactor behaves in a translational and rotational motion during the collision with a vehicle. The behavior of the impactor during the crash event was captured by a high speed camera and is regarded as the four-degree-of-freedom system in terms of translational and rotational motions. Pedestrian injuries are evaluated by three physical quantities indexes: the acceleration of the tibia, both the displacement and the bending angle between the femur and the tibia. The physical model for the impactor was expressed mathematically by differential equations. In the case of modelling, the ligament connecting both the femur and the tibia in particular plays an important role. Shear forces were applied to the ligament in translational motions and the bending moments in rotational motion. Differential equations were expressed in the form of a state equation and an output equation by MATLAB. Numerical solutions were obtained by a block diagram with Simulink. As a result, it was found that the predicted injuries agree quite well with their experimented data in terms of acceleration, displacement, and the bending angle mentioned above.

Pedestrian Crash Prediction Models and Validation of Effective Factors on Their Safety (Case Study: Tehran Signalized Intersections)

The quantity and severity of traffic accidents have increased with the development of machinery life and traffic growth in cities and roads in the past 50 years. Among the road users, pedestrians are the most vulnerable groups to be exposed to high risks. Vehicle crashes with pedestrian are almost inevitable and cause injury or death to pedestrian. Crash investigation and statistical studies indicate that percentage of pedestrian deaths caused by vehicle accidents are much more than all deaths. A considerable amount of accidents occur at signalized and urban intersections which are the intensive crash places. Therefore in this paper appropriate models that could specify safety indicators have been indicated with existing information by characterized parametric and nonparametric variables for twenty signalized intersections. Categories and correlations of variables also have been investigated. Three models including Regression, Poisson, and Negative binomial with defined variables have been determined. T and chi square tests, calibration and comparison of variables have been done by curve fitting. The role of each parameter was specified in pedestrian crashes. Validating models had the following outcomes: Pedestrian crash prediction models were based on none linear relations at intersections. Predictable variables, developing extended linear models and also pedestrian crash prediction are on the basis of Negative binomial distribution which is used due to more data dispersion. As observed, the Negative binomial regression because of its more R2 correlation factor has more validity among other regression models such as linear regression and Poisson. Calibrated models are put into sensitivity analysis to study the effect of each previously mentioned parameter in overall performance. Hence much better perception of future transportation plans can be achieved by development of safety models at planning levels.

Pedestrian Detection Method Based on SSD Model

Pedestrian detection has a wide range of applications in daily life, and many fields require pedestrians to conduct detection with high precision and speed, which is an urgent problem to be solved. The traditional pedestrian detection method improves the detection performance by improving the classification algorithm and extracting more effective features. In this paper, a pedestrian detection method is proposed based on single shot multibox detector (SSD) model, which replaces the basic network part of SSD model with inception network structure with smaller parameters, faster running speed and stronger nonlinear expression ability. A high-performance network model for pedestrian detection was based on improved SSD. The experimental results show that the proposed method is faster than the original model, and the average precision of pedestrian recognition and location is 89.6%, which is 2.6% higher than the original model.

车人碰撞中人地接触损伤仿真及防护方法鲁棒性分析

人车事故中行人主要与车辆和地面接触而遭受损伤。此前已有大量研究在车辆接触阶段对多体行人模型进行了验证,但对地面接触阶段的验证较少。本文中评估了用于预测车辆撞击后行人运动和地面接触的4个PC-Crash行人模型。通过车辆和地面接触的HIC伤害与最近研究中的6个尸体实验数据对比显示,其中PC2014行人模型能很好地预测行人头部损伤,车辆和地面接触HIC的平均误差分别为6.07%和5.85%,数据说明PC2014行人模型可以用于再现人车碰撞事故以及开发未来地面接触伤害防护对策。控制制动防护方法鲁棒性研究显示,在3种行人姿态(奔跑、步行和应急)扰动下,HIC平均降低比例分别为63.2%、57.9%和67.8%,说明控制制动防护方法具有很好的抗行人姿态干扰能力。进一步分析不同姿态下的行人损伤发现,3种步态序列之间均有显著性差异,在应急步态下控制制动防护方法有着更好的抗行人姿态干扰能力;而在奔跑步态序列下,往往可以通过控制制动防护方法产生最低的头地碰撞损伤,表明在使用控制制动防护方法对行人进行保护时须考虑到不同姿态产生的影响,以进一步提高控制制动防护方法的地面伤防护效果。

行人过街模拟及车辆右转避障路径规划方法

为解决无信号十字路口右转车辆与同侧过街行人的交互冲突问题,提出一种模拟过街行为的行人过街运动模型,设计了车辆横纵向解耦避障路径规划算法,并进行了仿真实验。使车辆面向动、静态行人时能合理切换避障路径规划策略;同时,将过街运动模型驱动下的行人作为车辆避障对象,以过街模型输出的行人未来轨迹生成车辆纵向速度规划障碍位移—时间区域,从而让行人未来运动状态反馈到车辆避障中。结果表明:本文的行人过街运动模型相对观测值的准确率达到了90%,因此,该模型复现了行人过街过程;能根据行人运动状态切换避障方案,使车辆安全避让过街行人。

基于AnyLogic仿真的恐慌情景行人疏散研究

为了反映恐慌情景下行人疏散的变化规律,基于Anylogic软件进行仿真分析。通过引入情绪因子和信息因子,量化恐慌传播系数和信息传播系数,建立群体恐慌动力学模型,运用蒙特卡洛等方法进行群体心理的量化分析,划分恐慌程度等级为Ⅰ级、Ⅱ级、Ⅲ级、Ⅳ级、Ⅴ级。运用该模型分析不同密度、不同恐慌等级下的恐慌人员疏散规律。研究结果表明:低密度情景下恐慌对疏散效率的影响较低,中、高密度情景下,适度恐慌可以降低疏散时间,超出临界范围会导致疏散的延缓。场景模拟可以看出,适度恐慌下人员疏散迅速有序,相反,过度恐慌下人员疏散缓慢混乱。