Identification And Evaluation Of Kayak Technical Movements Based On Wearable Computing

Kayak is a very popular outdoor sport,with the development of sports cause,kayak is gradually known by the majority of enthusiasts in China.However,kayak sports start late in our country and have a big gap with the world’s top teams.Moreover,after the 2008 Beijing Olympic Games,due to lack of reserve talents,the kayak cause once fell into a low period.In recent years,along with the concepts of Tech-Sport proposed,motion capture technology has been widely used in the field of sports,to assist athletes training,improve the performance and reduce the risk of injury.As an aquatic sport,kayak has a wide range of activities and complex environment.Therefore,the current research on kayak is still lacking of effective monitoring means.Based on previous studies,this paper proposes a motion capture scheme for kayak.The main research contents are as follows:(1)Based on inertial measurement unit(IMU),this paper designs a wearable motion capture system for the technical movement monitoring of kayak,so as to realize the collection,analysis and quantitative evaluation of athletes’ motion data.In order to limit the error of gyroscope attitude estimation,gradient descent method is used for multi-sensor data fusion to realize the updating of athletes’ posture,and a method of quaternion driven human skeleton vector model is proposed to reconstruct the athletes’ rowing movements.The angle sequence of left shoulder,right shoulder,left elbow and right elbow joint of upper limb is calculated and compared with the optical motion capture system.The results show that the motion capture system developed in this paper is comparable to the optical motion capture system in the term of measurement accuracy.(2)Aiming at the problem of laborious manual phase labeling in traditional video analysis method,an automatic phase segmentation method for kayak rowing is proposed combined with machine learning algorithm.According to the permutation and combination principle,the athlete’s upper limb joint angle sequence is combined in 9 different ways to study the influence of different combination ways on the accuracy of phase segmentation of four machine learning algorithms(decision tree,SVM,KNN,Bagging).The results indicating that the combination of shoulder angle sequence and SVM algorithm could better complete the task of the phase segmentation for kayak rowing.The classification accuracy of this method can reach 98.1%,only three data acquisition nodes are needed,and the calculation cost is small.(3)On the basis of phase segmentation,the detailed analysis of kayak athletes’ rowing rhythm,limbs and waist technical movements and rowing stability is realized.Referring to the average data of male kayak athletes of the national team,the competitive ability of kayakers is evaluated quantitatively.The results show that the wearable motion capture system designed in this study can provide more reference information for coaches in the stage of technical diagnosis and competitive ability evaluation,and provide a new idea for kinematics monitoring of kayak.