Research On Moving Object Detection And Estimation Algorithm Of Table Tennis Automatic Scorer

In sports competitions at home and abroad,the referees often lead to a few miscarriages of justice.The table tennis known as national ball in China is not exceptive.It mainly relies on manual referee and often brings dispute especially in respect of rubbing and wiping.On the one hand,as table tennis is so fast that the speed can reach 17m/s,rubbing and wiping the net is only an instant process.However,human vision has a temporary phenomenon,and the highest frequency to distinguish things is 24 Hz.It takes at least 0.042 s to respond,which makes it difficult for them to distinguish between some subtle differences in the moment.Therefore,it is difficult to rely on artificial judgments just to ensure the accuracy of the referee and it needs more detailed basis for the referee.On the other hand,the artificial decision is time-consuming and laborious.It is urgent for us to develop an intelligent device to help the referees making their decisions by offering reliable data.In the current ball automatic referee,there are some problems such as high cost of equipment,processing of large data and poor real-time problem.In view of the above problems,we design a high performance and high efficiency moving object detection algorithm and motion estimation algorithm.In the aspect of moving object detection,we design an efficient target detection algorithm,and sets reasonable detection conditions,which can extract ping-pong from real-time video.In the aspect of motion estimation,We study the motion mechanism of table tennis,designs the motion estimation algorithm with high real-time and high accuracy,and solves the problem of lack of moving target position information in the image.The simulation results show that the error of motion estimation is less than 3.2cm.Compared with the 4.0 cm diameter of table tennis,the algorithm provides reliable data for accurate referee.In addition,in order to explore the effect of Magnus force on the trajectory of the sensory cognition,we overcome the problem that the motion estimation of rotating table tennis stays in the sensory cognition,which provides an important theoretical basis for the research on the motion estimation.In order to solve the problem of missing a large number of moving objects in the video,we build the mode according to the complex motion mechanism and solves the differential equation to get the position information.The simulation results show that the error of moving target missing is less than 4 cm,which is smaller than the size of table tennis table and the movement trend is consistent with the actual movement trend.