The Key Technology Of The Parrot Earthquake Behavior Based On Computer Vision Research

In recent years,frequent earthquakes around the world caused huge casualties and property losses.However,the current earthquake prediction is still a worldwide problem.There are many legends about abnormal animals before the earthquake in ancient China.In recent years,Japan and the United States also have some reports anomalies about animals on earthquake,but these are some of the folk records and speculation,there is no scientific evidence to show the exact relationships.In order to illustrate if there was a connection,the paper selected many experts believe that prior to the earthquake response obviously budgies as the research object,analyzed its associated behaviors.This thesis summarized and analyzed the current situation of animal behavior research analysis,including the technical methods used and problems,and for the research object-parrots,using the method based on computer vision,parrot impending earthquake behavior of key technologies the study.Details are as follows:1.The use of background difference algorithms extracts the target and the algorithm is improved in response to changing lighting and environmental noise adaptive background updating,you can quickly extract stationary and motion.2.Form objectives were analyzed,using the global features and local features a combination of methods and the use of machine learning methods for several key morphological objectives are classified.3.The use of continuous-time changes in the shape and position of the target to identify the target of beating behavior,and take the form of customized regional situation and the method of combining parrot,parrot eating behavior analysis.4.The statistics out various acts of law parrot behavior was analyzed using the frequency of parties activity parrot beating of law analysis,the use of Cartesian product method combines time factor,environmental factors,behavioral factors and health factors combine parrot behavior coding,combined with parrot track density distribution,find the law of distribution of the target track.Finally,the use of java language to implement the automatic analysis parrot behavior and record-keeping systems for a wide range of real-world data to support related research.Finally,the article points out the current problems of the next research work in the future.