CAJ | 학술논문

对肥胖人群运动特征不合理性进行视觉图像分割,排除不合理的运动要素成分,算法的设计是实现人体运动健康监测仪的核心软件.传统方法中对人体运动特征的视觉图像分割采用全局同态分割方法,不适合对肥胖人群运动特征不合理性的提取.提出一种基于局部特征匹配的肥胖人群运动特征不合理性的视觉图像分割算法.对肥胖人群的运动特征图像区域特征进行边缘结构层分析,采用局部特征匹配方法实现对不合理的运动特征部分的分割处理.仿真表明,采用该方法能实现对肥胖人群运动特征不合理性信息要素有效滤除,提高视觉图像分割的准确性,实现运动状态的监测和视觉评估.
대비반인군운동특정불합이성진행시각도상분할,배제불합리적운동요소성분,산법적설계시실현인체운동건강감측의적핵심연건.전통방법중대인체운동특정적시각도상분할채용전국동태분할방법,불괄합대비반인군운동특정불합이성적제취.제출일충기우국부특정필배적비반인군운동특정불합이성적시각도상분할산법.대비반인군적운동특정도상구역특정진행변연결구층분석,채용국부특정필배방법실현대불합리적운동특정부분적분할처리.방진표명,채용해방법능실현대비반인군운동특정불합이성신식요소유효려제,제고시각도상분할적준학성,실현운동상태적감측화시각평고.
The irrationality of obesity crowd motion characteristics of visual image segmentation, eliminate unreasonable fac-tors movement components, algorithm design is the implementation of core software of human motion health monitor. In the traditional method, the global image segmentation method is not suitable for the extraction of the characteristics of the mo-tion of the obese people. Proposed a visual image segmentation algorithm based on local feature matching for the motion of the obese crowd. Structure of edge layer for obese people with the motion characteristics of image features, motion character-istics of unreasonable were mutation and smooth region distinction, to the local feature matching method realization on the part of the movement characteristics of unreasonable segmentation. Simulation results show that using the method can achieve on obesity crowd motion characteristics of irrational elements effectively filter and improve the accuracy of the visu-al image segmentation, motion state monitoring and visual assessment.