科技通报
科技通報
과기통보
BULLETIN OF SCIENCE AND TECHNOLOGY
2015年
2期
58-60
,共3页
模板叠加%相异特征%CT图像%断点标定
模闆疊加%相異特徵%CT圖像%斷點標定
모판첩가%상이특정%CT도상%단점표정
template superposition%different features%CT image%breakpoint calibration
运动员因剧烈运动常导致隐性骨骼损伤,常规的CT图像观测方法靠人工经验判断,无法有效发现隐性骨骼损伤的CT图像断点,不能形成合理的诊断方案,断点标定性能较差。提出一种基于模板叠加和相异特征匹配的CT图像断点标定方法,进行基于小波变换的CT图像边缘检测分离,采用相异特征匹配的方法消除图像中的噪点信息,提取灰度像素值特征,根据Shannon信息论原理,描述融合结果与源图像的综合相关性信息,得到CT源图像与断点融合图像的进行相关性标定,采用模板叠加和相异特征匹配得到准确的断点标定结果。实验证明,采用该算法能有效实现对CT图像的隐性损伤骨骼损伤作出的断点标定,标定的准确度较高,鲁棒性好,抗噪性优越。为体育运动损伤诊疗提供理论依据。
運動員因劇烈運動常導緻隱性骨骼損傷,常規的CT圖像觀測方法靠人工經驗判斷,無法有效髮現隱性骨骼損傷的CT圖像斷點,不能形成閤理的診斷方案,斷點標定性能較差。提齣一種基于模闆疊加和相異特徵匹配的CT圖像斷點標定方法,進行基于小波變換的CT圖像邊緣檢測分離,採用相異特徵匹配的方法消除圖像中的譟點信息,提取灰度像素值特徵,根據Shannon信息論原理,描述融閤結果與源圖像的綜閤相關性信息,得到CT源圖像與斷點融閤圖像的進行相關性標定,採用模闆疊加和相異特徵匹配得到準確的斷點標定結果。實驗證明,採用該算法能有效實現對CT圖像的隱性損傷骨骼損傷作齣的斷點標定,標定的準確度較高,魯棒性好,抗譟性優越。為體育運動損傷診療提供理論依據。
운동원인극렬운동상도치은성골격손상,상규적CT도상관측방법고인공경험판단,무법유효발현은성골격손상적CT도상단점,불능형성합리적진단방안,단점표정성능교차。제출일충기우모판첩가화상이특정필배적CT도상단점표정방법,진행기우소파변환적CT도상변연검측분리,채용상이특정필배적방법소제도상중적조점신식,제취회도상소치특정,근거Shannon신식론원리,묘술융합결과여원도상적종합상관성신식,득도CT원도상여단점융합도상적진행상관성표정,채용모판첩가화상이특정필배득도준학적단점표정결과。실험증명,채용해산법능유효실현대CT도상적은성손상골격손상작출적단점표정,표정적준학도교고,로봉성호,항조성우월。위체육운동손상진료제공이론의거。
The athletes from strenuous exercise often leads to the recessive skeletal injury, the conventional method of CT image observations by artificial experience, cannot effectively find CT image breakpoint recessive skeletal injury diagnosis, reasonable scheme cannot be formed, it has the poor performance of the breakpoint calibration. A method is proposed to cal?ibrate the CT image matching template superimposed breakpoint and distinctive features based on separation, CT image edge detection based on wavelet transform, methods using different feature matching to eliminate image noise information, extracted from the gray level pixel values, according to the Shannon information theory, it describes the correlation with the source image information fusion results get the correlation, calibration and image of the breakpoints CT source image fu?sion, using template superposition and distinct feature matching to obtain accurate breakpoint calibration results. Experi?ments show that the algorithm can effectively achieve the CT image of recessive injury of skeletal injury of the breakpoint calibration, the calibration accuracy is improved, it has good robustness and excellent anti noise performance. It provides a theoretical basis for the diagnosis and treatment of sports injuries.