Carrying Capacity Mining And Tourism Emergency Prediction System Based On Swarm Intelligence Sensing

With the rapid development of national tourism in recent years,tourists are paying more attention to their self-interests and security during their travels.It is an important target of current tourism prediction study to use the related information of a scenic spot,which is collected,mined and analyzed,and videos to count people numbers and detect state in a scenic spot.Therefore,researches on carrying capacity mining and tourism emergency prediction system have quite important practical significance.The main work is completed as follows:(1)The thesis proposed an emotional comfort enhancement based algorithm(BCOF).The algorithm grabbed scenic spots related microblog data and comprehensively considered other scenic properties such as weather to predict current overall comfort conditions of visitors,accurately determined tourists' emotional tendencies and established a model of tourists' state expression.Experimental results show that the proposed BCOF algorithm gains a promotion of 14%in term of accuracy,and a promotion of 18%in term of recall rate,compared with BN and NEG algorithms,respectively.(2)The thesis used video surveillance to monitor the state of carrying capacity and proposed an anomaly detection algorithm based on social forces(SFNODE).On the basis of optical flow fused with social force model,the SFNODE algorithm analyzed different dimensions and used the pressure map of social force matrix to realize the detection of abnormal state in a scenic spot.Secondly,the thesis proposed a number density of statistical algorithm based on pressure(CTF-A).Based on the extracted scenic density,the CTF-A algorithm realized the prediction of people number in a scenic spot by establishing prediction function of people number under different density levels and transforming the density map of scenic area.Experimental results show that the proposed SFNODE algorithm gains a promotion of 9%and 2%in term of accuracy,compared with FLOW and Social Force algorithms respectively.And the proposed CTF-A algorithm gains a promotion of 5%in term of accuracy,compared with density-independent algorithm.(3)The thesis established a warning model of tourism emergency based on swarm intelligence sensing,dynamically detected scenic state,developed and implemented the corresponding verification systems.It fused comfort state mined from social network and tourist state from surveillance video with the index of early-warning model,and established a three-layer BP neural network.Experimental results show that the proposed prediction model improves the matching degree of the model and performs a higher prediction accuracy.(4)The thesis designed and developed a carrying capacity mining and tourism emergency prediction system based on swarm intelligence sensing.The system verified the effectiveness of the proposed BCOF,SFNODE and CTF-A algorithms.The verification results show that these proposed algorithms have good performances and can provide accurate information for scenic detection.