Direct generation of rough cut numerical control code from industrial CT slice images

This paper proposes a new method of generating a rough cut Numerical Control(NC) codes directly from industrial computerized tomography(CT) slice images for the products which work-piece did not need to be redesigned.The contour structure data of the work-piece are obtained by the threshold segmentation,edge extraction and contour tracing. The inner and outer contours are judged,and the tool radius compensation and the allowance reservation are performed by offsetting the contours according to the judgments. The authors generate the rough cut NC code cut area distinguishing and rough cut path generating in a slice by slice manner. In contour offsetting step, a new approach is proposed to judge inner and outer contours. A software system is developed to generate NC codes. An example is used to verify this approach and results show the feasibility and validity of the approach.     

Development and Application of a Ultrasound Image Management System

Aimed to enhance data processing ability and to server perfect information management of B-scan this system is emerged.Integration of with a standard PC and a B-scan with a video capture card,this system researched and developed a B-scan diagnostic and management system including image capture,image processing and data management function.The system implement image processing and analysis functions consisting of image enhancement,geometrical parameters calculation and so on.Otherwise it provides convenient diagnostic report creation function,powerful archive management and printing function,namely it realizes a efficient and cute PACS system.By incomparable advantage of standard PC,this system improves and upgrades the performance of traditional B-scan, improves the doctors' work efficiency greatly.     

结合植被指数与纹理特征的玉米冠层FAPAR遥感估算研究

光合有效辐射吸收比率（fraction of absorbed photosynthetically active radiation,FAPAR）是反映作物产量的重要参数之一。无人机遥感能够快速无损地获取高分辨率植被冠层光谱信息,已成为进行物理化参数反演的重要手段。以不同播期玉米为研究对象,基于无人机搭载多光谱传感器,提取植被指数与植被纹理特征,使用偏最小二乘（partial least squares regression,PLSR）方法将二者结合反演玉米FAPAR,并与传统单独使用植被指数或植被纹理特征反演植被FAPAR的方法进行比较。结果表明：使用传统方法单独利用植被指数反演FAPAR（验证RMSE最低为7.33×10^-2,rRMSE最低为8.66%）的效果比单独利用纹理特征反演FAPAR（验证RMSE最低为9.50×10^-2,rRMSE最低为11.23%）的精度更高;使用PLSR方法单独利用植被指数或纹理特征估算FAPAR的效果比传统方法精度更高（植被指数与纹理特征的验证RMSE最低分别为6.77×10^-2和5.24×10^-2,rRMSE最低分别为8.01%和6.19%）;使用PLSR方法将植被指数与纹理特征相结合估算FAPAR（验证RMSE最低为4.72×10^-2,rRMSE最低为5.57%）的效果比单独使用植被指数或纹理特征估算FAPAR的精度更高。综上,使用PLSR方法将植被指数和植被纹理特征相结合来反演玉米冠层FAPAR可行,为作物FAPAR遥感反演研究提供了新的思路。     

基于高分辨率影像的农村承包地调查研究

遥感影像地物信息丰富,这使得直接通过影像获取地块信息成为可能,因此借助高分辨率遥感影像数据,利用人机交互方式进行农村承包地利用状况调查的实践。介绍了应用高分辨率遥感影像数据获取农村承包地地块面积、地类和权属名称等信息的基本技术流程和方法,结论表明利用高分辨率遥感影像进行农村土地利用状况调查取得丁较好的效果,证明该方法具有较强的实用性。     

一株鳙源鲁氏耶尔森氏菌的分离鉴定及其感染的肠道病理损伤

为确定患病鳙鱼致病菌及筛选敏感药物,通过平皿划线法从患病鳙鱼血液及体表溃烂组织中分离到1株革兰氏阴性的短杆菌JL9510,该菌株在电镜下呈短杆状,在LB固体培养基上呈透明状菌落。VITEK^?MS全自动快速微生物质谱检测结果证明,该菌株为鲁氏耶尔森氏菌的置信度为99.9。通过16S rRNA及系统发育树构建分析,发现菌株JL9510与已知菌株JQ657818.1 Yersinia ruckeri聚为一支,确定其是一株鲁氏耶尔森菌。人工回归感染试验结果证明,该菌株为患病鳙鱼的致病菌。VITEK 2 Compact自动化细菌鉴定和药物敏感性分析系统结果显示,鲁氏耶尔森氏菌JL9510有较高的致病力,且对氟苯尼考敏感,对环丙沙星耐药。试验的抗生素表现出良好的治疗效果,为鲁氏耶尔森氏菌的防治奠定了基础。     

鱼类肝脏疾病及防治研究

作者于１９８６￣１９９６年间先后对国内外８个养鱼场的饲养鱼类肝病发生情况作了系统调查，通过体表检查、内脏解剖及肝组织病理观察，对鱼类肝病的症状及病型进行了分级和分类，并讨论了致病原因，提出了防治措施。     

基于1990-2020年综合土地利用动态变化的巩义市城市边缘区空间研究

以郑州市下辖且接壤的巩义市为研究对象,以1990年和2020年两期TM/OLI影像和统计年鉴数据为依据,使用1 km×1 km网格划分样区并提取各样区的综合土地利用动态度数值,借助数学模型分析确定城市边缘区范围,分析巩义市近30年来城市边缘区用地变化特征及扩展机制。结果表明,自然条件是用地扩张的基础因素,政策与经济是边缘区外扩的核心驱动力,交通对扩张具有指向性作用;边缘区内的土地变化主要表现为建设用地急剧增加、农业用地锐减,同时受距离和开发时序影响,用地类型交错复杂,破碎化特征明显。     

基于FE-P2Pnet的无人机小麦图像麦穗计数方法

针对无人机图像背景复杂、小麦密集、麦穗目标较小以及麦穗尺寸不一等问题,提出了一种基于FE-P2Pnet（Feature enhance-point to point）的无人机小麦图像麦穗自动计数方法。对无人机图像进行亮度和对比度增强,增大麦穗目标与背景之间的差异度,减少叶、秆等复杂背景因素的影响。引入了基于点标注的网络P2Pnet作为基线网络,以解决麦穗密集的问题。同时,针对麦穗目标小引起的特征信息较少的问题,在P2Pnet的主干网络VGG16中添加了Triplet模块,将C(通道)、H(高度)和W(宽度)3个维度的信息交互,使得主干网络可以提取更多与目标相关的特征信息;针对麦穗尺寸不一的问题,在FPN(Feature pyramid networks)上增加了FEM(Feature enhancement module)和SE(Squeeze excitation)模块,使得该模块能够更好地处理特征信息和融合多尺度信息;为了更好地对目标进行分类,使用Focal Loss损失函数代替交叉熵损失函数,该损失函数可以对背景和目标的特征信息进行不同的权重加权,进一步突出特征。实验结果表明,在本文所构建的无人机小麦图像数据集(Wheat-ZWF)上,麦穗计数的平均绝对误差（MAE）、均方误差（MSE）和平均精确度(ACC)分别达到3.77、5.13和90.87%,相较于其他目标计数回归方法如MCNN(Multi-column convolutional neural network)、CSRnet（Congested scene recognition network)和WHCNETs (Wheat head counting networks)等,表现最佳。与基线网络P2Pnet相比,MAE和MSE分别降低23.2%和16.6%,ACC提高2.67个百分点。为了进一步验证本文算法的有效性,对采集的其它4种不同品种的小麦（AK1009、AK1401、AK1706和YKM222）进行了实验,实验结果显示,麦穗计数MAE和MSE平均为5.10和6.17,ACC也达到89.69%,表明本文提出的模型具有较好的泛化性能。     

TRANSCUTANEOUS ULTRASONOGRAPHY OF THE AVIAN UROGENITAL TRACT

The purpose of this article is to provide a survey of the diagnostic use of transcutaneous sonographic examination of the urogenital tract of pet birds. Possibilities and limitations in imaging normal and altered components of this organ system are discussed. Ultrasonographic examination of the avian urogenital tract was evaluated in 386 pet birds. Abnormal clinical signs were present in 289 birds; 97 birds were normal. Transcutaneous ultrasonography proved to be a useful, noninvasive, and fast diagnostic tool, especially in imaging laminated eggs and changes of the oviduct. Until now, these pathologic changes were not diagnosable in birds with other common noninvasive methods (e.g. radiography). The imaging of low-grade changes of the oviduct, disorders of the kidney without enlargement of the organ, as well as roughening of the egg shell proved to be difficult. Sonographic imaging of neither the normal kidney nor the inactive gonads was possible.