基于图像处理的物体质量与软硬度的测量  

Measurement of mass and softness of an object based on image processing

作　　者：张学军[1,2,3] 瞿祥和 雷锦添 毕红梅[1] 刘天鹏[1] 

机构地区：[1]广西大学计算机与电子信息学院,广西南宁530004 [2]广西多媒体通信与网络技术重点实验室培育基地,广西南宁530004 [3]广西高校多媒体通信与信息处理重点实验室,广西南宁530004

出　　处：《广西大学学报（自然科学版）》2016年第2期541-553,共13页Journal of Guangxi University（Natural Science Edition）

基　　金：国家自然基金资助资助项目(81460274;61262027);广西大学大学生实验技能和科技创新能力训练基金资助项目(SYJN20130728)

摘　　要：物体的质量及软硬度的识别一般采用物理的方法来测量,操作复杂并且需要特定的工具,而对于物体内部则无法直接测量。随着机器人和计算机视觉技术的发展,利用图像处理的手段识别和分析物体的物理特性的研究成为可能。针对基于网格状的形变进行图像处理分析识别物体软硬度:在空间领域,采用薄板样条(TPS)算法对形变的最小弯曲能量值进行计算,对物体的柔软度进行量化分析;在频谱领域,采用傅里叶变换(FFT)算法计算出变形前后特征区域的频谱的变化值,寻找其与柔软度的关系曲线;将空间、频谱领域的实验结果整合分析,确定物体硬度的图像测定参数。物体所受作用力越大,产生的弯曲能量值越大,频谱扩散的效果越明显。针对一些柔软度较好的常见物体图像的实验效果进行对比分析,发现拉伸带的实验效果较好。对其施加不同大小的作用力,形变效果显著,且结果呈线性分布,符合理论预期。进而使用该测量方法对一组医学肝脏图像进行了分析,计量不同患者肝脏的柔软度,以帮助医生及时准确地诊断出肝纤维化程度,达到预防肝硬变及癌变的目的。In general,the mass and hardness of an object are measured by using physical method,which either is complex in operation or requires specific tools. Furthermore,it is not possible to directly measure the properties inside an object. With the rapid development of robots and computervision technology,image processing method makes it possible to identify and analyze the physical properties of an object. In this paper,we propose a novel method based on image processing to measure the hardness and softness of an object by analysis of the deformation degree in grid pattern：in the spatial field,the thin plate spline（ TPS） algorithm is adopted to calculate the minimum deformation bending energy,leading to a quantitative analysis method for object＇s softness; in the field of spectrum,the Fourier transform（ FFT） algorithm is used to calculate the difference of spectrum within a deformation frequency area before and after the change of grids,from which the relationship between softness and spectrum is investigated. Finally the experimental results are integrated in the field of space and spectrum to determine the measurement parameters of the hardness to the image.The results show that the more force on an object,the greater value of bending energy,and the greater spectrum spreading effect. Different soft materials are applied in our experiment for comparing and analyzing the experimental effect,we found that tensile has the best outcome to express the stiffness in image findings. By pressing different forces on this material,its deformation effect is remarkable,and the result is in a linear distribution as meeting the theoretical expectations. Our method of hardness measurement is applied to a set of medical images on liver with grid tags. The result indicates the ability of method on measuring the liver softness in different patients. It is expected that our method could help the doctor to diagnose the degree of liver fibrosis rapidly and accurately,and reach the goal of prevention of liver cirrhosis and cance

关 键 词：薄板样条 傅里叶变换 软硬度 频谱 弯曲能量 图像处理 

分 类 号：TN98[电子电信—信息与通信工程]