基于运动学数据采集和分析，明确了速度滑冰出发阶段的蹬冰角。采用4台1920*1080的高清摄像机拍摄运动员速度滑冰出发阶段的技术动作。每一大组（4台摄像机）分别为起始点2台一组、结束点2台一组，每组摄像机的主光轴夹角约为100°，拍摄频率60幅/s。使用Fastmove 3D Motion（锐动科技，大连，中国）的人工智能人体关节点自动识别功能对录像进行解析，并使用Butterworth 4阶低通滤波方法对蹬冰角结果进行低通滤波，截止频率为10.0Hz。同时，使用Matlab R2022a对时间进行归一化处理，得到了速度滑冰出发阶段的蹬冰角。

Based on kinematic data collection and analysis, the stirrup angle of the speed skating departure phase was clarified. Four 1920*1080 high-definition cameras were used to capture the technical movements of the athletes in the speed skating departure phase. Each large group (4 cameras) consisted of 2 cameras at the start point and 2 cameras at the end point, each with a main optical axis angle of about 100° and a shooting frequency of 60 frames/s. The video was analyzed using the artificial intelligence human joint point automatic recognition function of Fastmove 3D Motion (Dalian, China), and the results of the stance angle were analyzed using the Butterworth 4th order low-pass filtering method. The results were low-pass filtered with a cut-off frequency of 10.0 Hz using the Butterworth 4th-order low-pass filtering method, and normalized in time using Matlab R2022a to obtain the speed skating start angle.