期刊文献+
共找到17,761篇文章
< 1 2 250 >
每页显示 20 50 100
New multi-DSP parallel computing architecture for real-time image processing 被引量:4
1
作者 Hu Junhong Zhang Tianxu Jiang Haoyang 《Journal of Systems Engineering and Electronics》 SCIE EI CSCD 2006年第4期883-889,共7页
The flexibility of traditional image processing system is limited because those system are designed for specific applications. In this paper, a new TMS320C64x-based multi-DSP parallel computing architecture is present... The flexibility of traditional image processing system is limited because those system are designed for specific applications. In this paper, a new TMS320C64x-based multi-DSP parallel computing architecture is presented. It has many promising characteristics such as powerful computing capability, broad I/O bandwidth, topology flexibility, and expansibility. The parallel system performance is evaluated by practical experiment. 展开更多
关键词 parallel computing image processing REAL-TIME computer architecture
在线阅读 下载PDF
Dynamic experimental study on rock meso-cracks growth by digital image processing technique 被引量:2
2
作者 朱珍德 倪骁慧 +3 位作者 王伟 李双蓓 赵杰 武沂泉 《Journal of Central South University》 SCIE EI CAS 2008年第S2期114-120,共7页
A new meso-mechanical testing scheme based on SEM was developed to carry out the experiment of microfracturing process of rocks. The microfracturing process of the pre-crack marble sample on surrounding rock in the im... A new meso-mechanical testing scheme based on SEM was developed to carry out the experiment of microfracturing process of rocks. The microfracturing process of the pre-crack marble sample on surrounding rock in the immerged Long-big tunnel in Jinping Cascade II Hydropower Station under uniaxial compression was recorded by using the testing scheme. According to the stereology theory, the propagation and coalescent of cracks at meso-scale were quantitatively investigated with digital technology. Therefore, the basic geometric information of rock microcracks such as area, angle, length, width, perimeter, was obtained from binary images after segmentation. The failure mechanism of specimen under uniaxial compression with the quantitative information was studied from macro and microscopic point of view. The results show that the image of microfracturing process of the specimen can be observed and recorded digitally. During the damage of the specimen, the distribution of microcracks in the specimen is still subjected to exponential distribution with some microcracks concentrated in certain regions. Finally, the change law of the fractal dimension of the local element in marble sample under different external load conditions is obtained by means of the statistical calculation of the fractal dimension. 展开更多
关键词 MARBLE ROCK failure process digital image processing growth of meso-crack DYNAMIC OBSERVATION experimental study
在线阅读 下载PDF
Image processing algorithm acceleration using reconfigurable macro processor model 被引量:2
3
作者 SunGuanKfu ChenHuaming LuHuanzhang 《Journal of Systems Engineering and Electronics》 SCIE EI CSCD 2004年第2期110-114,共5页
The concept and advantage of reconfigurable technology is introduced. A kind of processor architecture of re configurable macro processor (RMP) model based on FPGA array and DSP is put forward and has been implemented... The concept and advantage of reconfigurable technology is introduced. A kind of processor architecture of re configurable macro processor (RMP) model based on FPGA array and DSP is put forward and has been implemented. Two image algorithms are developed: template-based automatic target recognition and zone labeling. One is estimating for motion direction in the infrared image background, another is line picking-up algorithm based on image zone labeling and phase grouping technique. It is a kind of 'hardware' function that can be called by the DSP in high-level algorithm. It is also a kind of hardware algorithm of the DSP. The results of experiments show the reconfigurable computing technology based on RMP is an ideal accelerating means to deal with the high-speed image processing tasks. High real time performance is obtained in our two applications on RMP. 展开更多
关键词 real-time image processing reconfigurable computing technology reconfigurable macro processor model template matching image zone labeling.
在线阅读 下载PDF
Pore structure of ore granular media by computerized tomography image processing 被引量:6
4
作者 吴爱祥 杨保华 +1 位作者 习泳 江怀春 《Journal of Central South University of Technology》 EI 2007年第2期220-224,共5页
The pore structure images of ore particles located at different heights of leaching column were scanned with X-ray computerized tomography (CT) scanner, the porosity and pore size distribution were calculated and the ... The pore structure images of ore particles located at different heights of leaching column were scanned with X-ray computerized tomography (CT) scanner, the porosity and pore size distribution were calculated and the geometrical shape and connectivity of pores were analyzed based on image process method, and the three dimensional reconstruction of pore structure images was realized. The results show that the porosity of ore particles bed in leaching column is 42.92%, 41.72%, 39.34% at top, middle and bottom zone, respectively. Obviously it has spatial variability and decreases appreciably along the height of the column. The overall average porosity obtained by image processing is 41.33% while the porosity gotten from general measurement method in laboratory is 42.77% showing the results of both methods are consistent well. The pore structure of ore granular media is characterized as a dynamical space network composed of interconnected pore bodies and pore throats. The ratio of throats with equivalent diameter less than 1.91 mm to the total pores is 29.31%, and that of the large pores with equivalent diameter more than 5.73 mm is 2.90%. 展开更多
关键词 ore granular media pore structure X-ray computerized tomography image processing
在线阅读 下载PDF
An Algorithm for Ship Wake Detection from the SAR Images Using the Radon Transform and Morphological Image Processing 被引量:2
5
作者 金亚秋 王世庆 《Journal of Systems Engineering and Electronics》 SCIE EI CSCD 2001年第4期7-12,共6页
Using the Radon transform and morphological image processing, an algorithm for ship's wake detection in the SAR (synthetic aperture radar) image is developed. Being manipulated in the Radon space to invert the gra... Using the Radon transform and morphological image processing, an algorithm for ship's wake detection in the SAR (synthetic aperture radar) image is developed. Being manipulated in the Radon space to invert the gray-level and binary images, the linear texture of ship wake in oceanic clutter can be well detected. It has been applied to the automatic detection of a moving ship from the SEASAT SAR image. The results show that this algorithm is well robust in a strong noisy background and is not very sensitive to the threshold parameter and the working window size. 展开更多
关键词 ALGORITHMS image processing Mathematical transformations Radar clutter Radar target recognition Spurious signal noise Synthetic aperture radar
在线阅读 下载PDF
3D surface profile diagnosis using digital image processing for laboratory use 被引量:2
6
作者 Robert FRISCHER Ondrej KREJCAR +1 位作者 Ali SELAMAT Kamil KUCA 《Journal of Central South University》 SCIE EI CAS CSCD 2020年第3期811-823,共13页
The measurement of the surface quality and the profile preciseness is major issues in many industrial branches such that the surface quality of semi products directly affects the subsequent production steps.Although,t... The measurement of the surface quality and the profile preciseness is major issues in many industrial branches such that the surface quality of semi products directly affects the subsequent production steps.Although,there are many ways to obtain required data,the hardware necessary for the measurements such as 2D or 3D scanners,depending on the problem’s complexity,is too expensive.Therefore,in this paper,what we put forward as a novelty is an algorithm which is verified on the model of simple 3D scanner on the image processing basis with the resolution of 0.1 mm.There are many ways to scan surface profile;however,the image processing currently is the most trending topic in industry automation.Most importantly,in order to obtain surface images,standard high resolution reflex camera is used and thus the post processing could be realized with MatLab as the software environment.Therefore,this solution is an alternative to the expensive scanners,and single-purpose devices could be extended by many additional functions. 展开更多
关键词 profile diagnostics image processing 3D surface MatLab measurement
在线阅读 下载PDF
Numerical simulation of direct shear tests on mechanical properties of talus deposits based on self-adaptive PCNN digital image processing 被引量:5
7
作者 王盛年 徐卫亚 +1 位作者 石崇 张强 《Journal of Central South University》 SCIE EI CAS 2014年第7期2904-2914,共11页
The macro mechanical properties of materials with characteristics of large scale and complicated structural composition can be analyzed through its reconstructed meso-structures.In this work,the meso-structures of tal... The macro mechanical properties of materials with characteristics of large scale and complicated structural composition can be analyzed through its reconstructed meso-structures.In this work,the meso-structures of talus deposits that widely exist in the hydro-power engineering in the southwest of China were first reconstructed by small particles according to the in-situ photographs based on the self-adaptive PCNN digital image processing,and then numerical direct shear tests were carried out for studying the mechanical properties of talus deposits.Results indicate that the reconstructed meso-structures of talus deposits are more consistent with the actual situation because the self-adaptive PCNN digital image processing has a higher discrimination in the details of soil-rock segmentation.The existence and random distribution of rock blocks make the initial shear stiffness,the peak strength and the residual strength higher than those of the "pure soil" with particle size less than 1.25 cm apparently,but reduce the displacements required for the talus deposits reaching its peak shear strength.The increase of rock proportion causes a significant improvement in the internal friction angle of talus deposit,which to a certain degree leads to the characteristics of shear stress-displacement curves having a changing trend from the plastic strain softening deformation to the nonlinear strain hardening deformation,while an unconspicuous increase in cohesion.The uncertainty and heterogeneity of rock distributions cause the differences of rock proportion within shear zone,leading to a relatively strong fluctuation in peak strengths during the shear process,while movement features of rock blocks,such as translation,rotation and crossing,expand the scope of shear zone,increase the required shear force,and also directly lead to the misjudgment that the lower shear strength is obtained from the samples with high rock proportion.That,however,just explains the reason why the shear strength gained from a small amount of indoor test data is not consistent with engineering practice. 展开更多
关键词 talus deposits digital image processing pulse coupled neural networks(PCNN) direct shear test mechanical property granular discrete element method
在线阅读 下载PDF
Study on the Medical Image Distributed Dynamic Processing Method
8
作者 Zhang Quanhai & Shi PengfeiInstitute of Image Processing and Pattern Recognition, Shanghai Jiaotong University, Shanghai 200030, P. R. China 《Journal of Systems Engineering and Electronics》 SCIE EI CSCD 2003年第4期69-76,共8页
To meet the challenge of implementing rapidly advanced, time-consuming medical image processing algorithms, it is necessary to develop a medical image processing technology to process a 2D or 3D medical image dynamica... To meet the challenge of implementing rapidly advanced, time-consuming medical image processing algorithms, it is necessary to develop a medical image processing technology to process a 2D or 3D medical image dynamically on the web. But in a premier system, only static image processing can be provided with the limitation of web technology. The development of Java and CORBA (common object request broker architecture) overcomes the shortcoming of the web static application and makes the dynamic processing of medical images on the web available. To develop an open solution of distributed computing, we integrate the Java, and web with the CORBA and present a web-based medical image dynamic processing methed, which adopts Java technology as the language to program application and components of the web and utilies the CORBA architecture to cope with heterogeneous property of a complex distributed system. The method also provides a platform-independent, transparent processing architecture to implement the advanced image routines and enable users to access large dataset and resources according to the requirements of medical applications. The experiment in this paper shows that the medical image dynamic processing method implemented on the web by using Java and the CORBA is feasible. 展开更多
关键词 medical image dynamic processing based on web distributed computing INTEROPERATION CORBA.
在线阅读 下载PDF
High performance reconfigurable hardware system for real-time image processing
9
作者 赵广州 张天序 +2 位作者 王岳环 曹治国 左峥嵘 《Journal of Systems Engineering and Electronics》 SCIE EI CSCD 2005年第3期502-509,共8页
A novel reconfigurable hardware system which uses both muhi-DSP and FPGA to attain high performance and real-time image processing are presented. The system structure and working principle of mainly processing multi-B... A novel reconfigurable hardware system which uses both muhi-DSP and FPGA to attain high performance and real-time image processing are presented. The system structure and working principle of mainly processing multi-BSP board, extended multi-DSP board are analysed. The outstanding advantage is that the communication among different board components of this system is supported by high speed link ports & serial ports for increasing the system performance and computational power. Then the implementation of embedded real-time operating systems (RTOS) by us is discussed in detail. In this system, we adopt two kinds of parallel structures controlled by RTOS for parallel processing of algorithms. The experimental results show that exploitive period of the system is short, and maintenance convenient. Thus it is suitable for real-time image processing and can get satisfactory effect of image recognition. 展开更多
关键词 MULTI-DSP fidd programmable gate arrays real-time image processing real time operating systems parallel structure.
在线阅读 下载PDF
Automatic Detection of Steel Ball's Surface Flaws Based on Image Processing 被引量:1
10
作者 于正林 谭微 +1 位作者 杨东林 曹国华 《Defence Technology(防务技术)》 SCIE EI CAS 2007年第3期209-214,共6页
A new method to detect steel ball's surface flaws is presented based on computer techniques of image processing and pattern recognition. The steel ball's surface flaws is the primary factor causing bearing fai... A new method to detect steel ball's surface flaws is presented based on computer techniques of image processing and pattern recognition. The steel ball's surface flaws is the primary factor causing bearing failure. The high efficient and precision detections for the surface flaws of steel ball can be conducted by the presented method, including spot, abrasion, burn, scratch and crack, etc. The design of main components of the detecting system is described in detail including automatic feeding mechanism, automatic spreading mechanism of steel ball's surface, optical system of microscope, image acquisition system, image processing system. The whole automatic system is controlled by an industrial control computer, which can carry out the recognition of flaws of steel ball's surface effectively. 展开更多
关键词 计算机应用 钢球 表面裂痕 图像加工
在线阅读 下载PDF
人工智能在小肠息肉图像无创检测领域的研究进展 被引量:1
11
作者 张新峰 高子君 +1 位作者 刘晓民 李相生 《北京工业大学学报》 北大核心 2026年第2期148-157,共10页
小肠息肉起病隐匿,临床症状特异性不强,检出有一定难度,内窥镜检查技术是最常用的小肠疾病检查技术,但此技术操作复杂,亦有一定的观察盲区,如盲肠后方、肠瓣膜后方。通过计算机断层扫描(computed tomography,CT)、核磁共振(magnetic res... 小肠息肉起病隐匿,临床症状特异性不强,检出有一定难度,内窥镜检查技术是最常用的小肠疾病检查技术,但此技术操作复杂,亦有一定的观察盲区,如盲肠后方、肠瓣膜后方。通过计算机断层扫描(computed tomography,CT)、核磁共振(magnetic resonance imaging,MRI)等无盲区的非侵入式检测方式进行病变定位识别,具有重要临床意义,利用人工智能技术有望提高小肠息肉诊断的敏感性、准确性和快捷性。鉴于此,分析了人工智能技术在小肠息肉无创检测中的最新研究进展,内容包括:图像分割、小肠息肉三维重建、小肠息肉疾病分类预测。旨在助力提升小肠息肉检测和诊断的准确率;明晰技术发展脉络,为后续研究提供方向。 展开更多
关键词 小肠息肉 医学图像处理 深度学习 图像分割 三维重建
在线阅读 下载PDF
循环荷载下混凝土损伤断裂破坏过程研究
12
作者 郑丹 陈威 +1 位作者 赵斌 李鑫鑫 《重庆交通大学学报(自然科学版)》 北大核心 2026年第4期18-28,共11页
为探究循环荷载下混凝土裂缝的扩展规律与失效机理,采用数字图像相关(DIC)技术分析了荷载-裂纹张开位移曲线(P-CMOD)的演变过程,量化分析了裂缝口张开位移(CMOD)、裂缝长度a及断裂过程区残余应变随循环次数与卸载位置的变化特征。研究... 为探究循环荷载下混凝土裂缝的扩展规律与失效机理,采用数字图像相关(DIC)技术分析了荷载-裂纹张开位移曲线(P-CMOD)的演变过程,量化分析了裂缝口张开位移(CMOD)、裂缝长度a及断裂过程区残余应变随循环次数与卸载位置的变化特征。研究结果表明:在等幅循环荷载下,CMOD的演化呈现出快速累积、线性增长、失稳破坏等3个阶段特征。a与CMOD变化高度相关:在0.4F_(max)下,a呈线性缓慢增长后趋于稳定,随应力水平提高,a前期快速扩展,后期增速减缓,残余应力促进裂纹稳定扩展;在包络循环荷载下,a增速收敛而CMOD持续扩展,断裂过程区宽度增加,残余应变累积速率随荷载水平显著提高;在0.6F_(max)下,其应变增速为0.4F_(max)的2.6倍,高卸载残余应力加剧应变集中。基于分段力学响应与动态损伤演化机制,建立了可预测裂纹扩展速率、失效临界点及多荷载工况响应的P-CMOD曲线模型,为混凝土结构疲劳损伤评估与寿命预测提供基础。 展开更多
关键词 港口工程 残余应变 断裂过程区 三点弯曲实验 数字图像相关技术 循环荷载
在线阅读 下载PDF
我国海水鱼罐头加工与潜在质量问题剖析
13
作者 顾赛麒 蒋若萱 +3 位作者 周绪霞 丁玉庭 戴王力 徐心愉 《食品工业科技》 北大核心 2026年第11期487-497,共11页
海水鱼罐头是最常见的海产加工制品之一,具有营养丰富、风味多样、货架期长、食用便捷等优点,深受消费者欢迎。然而,我国海水鱼罐头产业起步较晚,当前在生产加工与贮藏环节仍存在一定的品质缺陷和食品安全风险,制约了行业的高质量发展... 海水鱼罐头是最常见的海产加工制品之一,具有营养丰富、风味多样、货架期长、食用便捷等优点,深受消费者欢迎。然而,我国海水鱼罐头产业起步较晚,当前在生产加工与贮藏环节仍存在一定的品质缺陷和食品安全风险,制约了行业的高质量发展。为此,本文以海水鱼罐头为研究对象,系统梳理了市售不同品牌产品的信息,分析其主要加工工艺及关键控制节点,归纳生产过程中常见的食品安全隐患,旨在为优化加工工艺、提升产品品质与安全性提供理论依据与技术参考。 展开更多
关键词 海水鱼 罐头 加工工艺 水产品品质 食品安全
在线阅读 下载PDF
MPCVD金刚石生长中的尺寸参数测量方法研究
14
作者 李洁 赵凤霞 +1 位作者 赵建国 张琳娜 《机械设计与制造》 北大核心 2026年第3期146-150,共5页
针对微波等离子化学气相沉积(MPCVD)制备的金刚石的形貌尺寸在线检测需求,研究了一种基于视觉图像分析的金刚石尺寸在线检测方法。采用工业相机在线采集金刚石生长图像后,首先在H分量图像上进行阈值分割、形态学处理、连通域分析等操作... 针对微波等离子化学气相沉积(MPCVD)制备的金刚石的形貌尺寸在线检测需求,研究了一种基于视觉图像分析的金刚石尺寸在线检测方法。采用工业相机在线采集金刚石生长图像后,首先在H分量图像上进行阈值分割、形态学处理、连通域分析等操作分离出每个金刚石图像;然后基于CLAHE方法增强每个分离出的金刚石图像,基于Canny算法提取出金刚石的内外边缘;最后对提取的金刚石边缘提出一种基于概率霍夫变换的尺寸测量方法,并对测得的尺寸进行了尺寸校正。实验验证结果表明,该方法能准确地实现金刚石轮廓尺寸的在线测量,长、宽、高的平均测量精度分别为98.727、99.071、97.999%,满足MPCVD金刚石的测量需求。 展开更多
关键词 MPCVD金刚石 参数测量 图像处理 相机标定 机器视觉
在线阅读 下载PDF
基于自监督学习的玉米植株图像小样本语义分割模型
15
作者 邓寒冰 刘鑫 +1 位作者 李朝阳 苗腾 《农业机械学报》 北大核心 2026年第1期72-82,共11页
图像语义分割技术是获取玉米植株表型信息的重要手段之一,传统的全监督语义分割方法往往依赖大量像素级标签,但玉米在不同生长阶段形态多变,导致图像标注成本高昂,制约模型在实际生产中的应用。为了去掉模型训练中的人工标注过程,本研... 图像语义分割技术是获取玉米植株表型信息的重要手段之一,传统的全监督语义分割方法往往依赖大量像素级标签,但玉米在不同生长阶段形态多变,导致图像标注成本高昂,制约模型在实际生产中的应用。为了去掉模型训练中的人工标注过程,本研究提出了一种基于自监督学习的玉米植株图像小样本语义分割网络(Self-supervised few-shot semantic segmentation network for maize plant images,MSDANet),以提高不同生长时期玉米植株图像的语义分割精度和模型泛化能力。MSDANet利用基于超像素的自监督学习方法生成伪标签,无需人工标注即可为支持集图像构建初步监督信号;设计混合遮蔽机制(Mixed masking,MM),应用基于伪标签的语义遮蔽,在特征空间构建多样性遮蔽样本,促进模型学习更鲁棒性的特征表达,从而提高复杂背景下的分割精度。针对图像中玉米植株存在的弯曲、重叠、遮挡等复杂形态问题,本研究为模型设计了多尺度可变形大核卷积注意力机制(Multi-scale deformable large kernel attention,MS-DLKA),通过融合多尺度感受野和可变形卷积,能够灵活感知玉米植株在不同尺度下的重要结构信息,有效提高了语义分割精度。在小样本数据集上进行验证,在1-shot设置下,MSDANet的mIoU和FB-IoU分别达到75.63%和87.12%;在5-shot设置下,mIoU和FB-IoU分别达到76.04%和87.21%,均优于本研究给出的同类其他模型。此外,与当前主流的全监督小样本语义分割模型对比,在1-shot和5-shot设置下,mIoU分别提升2.9、2.93个百分点。结果表明,MSDANet模型能够在无人工标签和小样本的前提下,实现高精度的玉米植株图像语义分割任务,为不同生长时期的玉米图像分析与植物表型测量提供了技术支持。 展开更多
关键词 玉米图像 植物表型 图像处理 深度学习 语义分割 自监督学习
在线阅读 下载PDF
ChestDS:基于深度学习的胸部低剂量CT降噪模型
16
作者 李子元 李建辉 +4 位作者 张晓东 刘佳 李玮 刘建新 王鹤 《中国医学影像技术》 北大核心 2026年第1期124-127,共4页
目的基于深度学习(DL)建立胸部低剂量CT(LDCT)图像降噪模型并观察其效果。方法基于Res Nte图像特征提取编码器、UNet网络图像生成解码器及VGG网络图像风格损失函数,利用AAPM数据训练集(18351对2D图像)和AAPM数据验证集(128对2D图像)数... 目的基于深度学习(DL)建立胸部低剂量CT(LDCT)图像降噪模型并观察其效果。方法基于Res Nte图像特征提取编码器、UNet网络图像生成解码器及VGG网络图像风格损失函数,利用AAPM数据训练集(18351对2D图像)和AAPM数据验证集(128对2D图像)数据生成Chest DS降噪模型;于视觉评估测试集(来自单中心的100例LDCT图像)分析观察者间主观评分的一致性,比较利用模型降噪前、后主观评分,并根据峰值信噪比(PSNR)及结构相似性(SSIM)于AAPM数据测试集(211对2D图像)对降噪及显示解剖结构效果进行客观评价。结果降噪前、后肺窗噪声评分组内相关系数(ICC)分别为0.88及0.96(P均<0.05),纵隔窗噪声评分ICC分别为0.50及0.86(P均<0.05);图像质量评分及关键解剖结构显示评分均显著提升(P均<0.05)。Chest DS模型在AAPM数据测试集的PSNR为49.05d B、SSIM为0.99。结论Chest DS模型用于LDCT图像可在通过降噪提高图像质量的同时降低辐射剂量。 展开更多
关键词 胸部 体层摄影术 X线计算机 辐射剂量 图像处理 计算机辅助 深度学习
在线阅读 下载PDF
基于NEWUOA的CRS叠加成像技术
17
作者 韩复兴 王源 +4 位作者 高正辉 常志邈 马飞 秦昊 尚浩 《吉林大学学报(地球科学版)》 北大核心 2026年第1期386-396,共11页
共反射面元(common reflection surface,CRS)叠加成像技术通过使用相邻共中心点道集信息提高地震数据信噪比,是处理低信噪比、低覆盖次数地震数据的关键技术。其三参数优化的搜索方法无论是经典的分步搜索还是同步的全局搜索,最终都需... 共反射面元(common reflection surface,CRS)叠加成像技术通过使用相邻共中心点道集信息提高地震数据信噪比,是处理低信噪比、低覆盖次数地震数据的关键技术。其三参数优化的搜索方法无论是经典的分步搜索还是同步的全局搜索,最终都需要局部优化算法提升精度,而传统的Nelder-Mead单纯形算法搜索方式简单,计算效率较低。本文采用一种无导数优化算法——新型无约束优化算法(new unconstrained optimization algorithm,NEWUOA)作为CRS叠加的精度提升算法。该算法通过动态构建二次插值模型,采用信赖域迭代策略,利用目标函数曲率信息提升计算效率。文中采用层状起伏模型和Marmousi2模型对该算法进行测试。结果显示,针对简单的层状起伏模型和复杂的Marmousi2模型,基于NEWUOA得到的相干值优化结果均普遍优于或等于基于Nelder-Mead单纯形算法得到的相干值优化结果,与Nelder-Mead单纯形算法相比,计算效率分别提升约48%和45%。这表明NEWUOA在保证CRS叠加优化精度的同时,能够显著提升计算效率,适合作为精度提升的计算工具。 展开更多
关键词 CRS叠加成像技术 三参数优化 NEWUOA 无导数优化 Nelder-Mead单纯形算法
在线阅读 下载PDF
深度学习驱动的碳纤维管壁缺陷 镜面红外检测无损定量评估方法
18
作者 杨阳 陈洪恩 +3 位作者 刘博凡 仝宗飞 裴翠祥 陈振茂 《西安交通大学学报》 北大核心 2026年第6期97-109,共13页
针对大型可展开天线碳纤维复合材料管壁缺陷传统检测方法存在检测距离近、效率低、智能化不足的问题,提出了一种基于图像处理和深度学习算法的碳纤维管脱粘缺陷镜面激光红外检测智能化定量评估方法。首先,基于Otsu算法进行检测区域自动... 针对大型可展开天线碳纤维复合材料管壁缺陷传统检测方法存在检测距离近、效率低、智能化不足的问题,提出了一种基于图像处理和深度学习算法的碳纤维管脱粘缺陷镜面激光红外检测智能化定量评估方法。首先,基于Otsu算法进行检测区域自动识别提取,并对提取后的图像序列进行主成分分析,有效抑制噪声干扰并提升缺陷特征的对比度。其次,提出了一种基于深度学习算法的缺陷智能化识别和分割算法,通过决策级融合策略实现缺陷区域的精准快速识别,并基于区域约束分割策略和逻辑或聚合算法实现缺陷形状的准确分割。最后,通过缺陷矫正与三维还原算法,实现了缺陷的三维可视化重构。对含内部脱粘缺陷的碳纤维管试件进行检测,结果表明:采用所提方法得到的缺陷定量评估结果与CT扫描结果高度一致,各个缺陷识别面积的最大误差为7.7%,平均误差为5.1%,均在工程误差允许范围内,验证了所提方法的准确性和有效性。该研究为可展开天线碳纤维管的远距离、高效和智能化无损定量评估提供了一定的指导意义。 展开更多
关键词 激光红外检测 碳纤维复合材料管 脱粘缺陷 深度学习 图像处理
在线阅读 下载PDF
基于改进YOLOv5s的水稻秸秆量识别检测方法
19
作者 葛宜元 孙修涵 +3 位作者 孟庆祥 梁秋艳 马浏轩 杜爽 《农机化研究》 北大核心 2026年第2期116-123,共8页
水稻秸秆腐解周期是衡量秸秆还田效果的重要指标。为提升水稻秸秆腐解效果,准确测量留于田间的秸秆量,将水稻收割后的田间秸秆状态分为铺放于田间的堆叠秸秆和根茎留于原位的留茬秸秆两种,并提出了一种基于改进YOLOv5s的水稻秸秆量识别... 水稻秸秆腐解周期是衡量秸秆还田效果的重要指标。为提升水稻秸秆腐解效果,准确测量留于田间的秸秆量,将水稻收割后的田间秸秆状态分为铺放于田间的堆叠秸秆和根茎留于原位的留茬秸秆两种,并提出了一种基于改进YOLOv5s的水稻秸秆量识别算法。采用色彩追踪算法结合HSV模型,对堆叠秸秆量进行检测;通过深度学习算法、OpenCV算法结合HSV模型和注意力模块,对留茬秸秆进行检测;引入CBAM注意力模块,同时通过非极大值抑制去除重叠锚定框,以消除重复计数的影响,从而实现对秸秆量的精准检测。通过训练模块对2174张水稻秸秆图片进行深度学习,识别精确率可达92.092%,召回率为96.144%,目标检测损失值为2.397%。实际田间秸秆检测时,正确检出率可达85.85%。改进后的算法可有效检出留茬秸秆并通过串口返回秸秆数量,为秸秆腐解剂的精量施放和还田模式的建立提供更为精确的数据支持。 展开更多
关键词 水稻秸秆量识别 改进YOLOv5s 注意力机制 图像处理 非极大值抑制
在线阅读 下载PDF
基于单帧连续编码条纹的半全局立体匹配测量方法
20
作者 杨成禹 黄晨雷 +2 位作者 张世达 林雪竹 李丽娟 《红外与激光工程》 北大核心 2026年第1期328-342,共15页
为了解决双目立体测量时,散斑结构光法受到投影质量和离散编码特性制约难以实现弱纹理区域精确匹配的问题,提出了一种基于单帧连续编码条纹的半全局立体匹配测量方法。在编码方面,设计了一种与半全局立体匹配路径相关的单帧连续编码条纹... 为了解决双目立体测量时,散斑结构光法受到投影质量和离散编码特性制约难以实现弱纹理区域精确匹配的问题,提出了一种基于单帧连续编码条纹的半全局立体匹配测量方法。在编码方面,设计了一种与半全局立体匹配路径相关的单帧连续编码条纹,实现对空间高效且鲁棒的特征编码。在算法方面,针对编码条纹改进了代价初始化方法,为立体匹配提供高质量的初始代价;同时引入自适应分区域并行计算和分层视差预测加速机制,提高了测量效率。实验结果表明:文中方法对距离1 m的弧形弱纹理样件实现了正确匹配率为92.77%,均方根误差为0.89mm的高分辨率精确测量。在0.5~4.0 m的深度范围内,平面测量误差在2.0 m处为1.54、4.0 m处为3.89mm,台阶高度测量在2.0 m处平均绝对误差优于1.25mm,均方根误差优于1.45mm,证明了该方法在较大范围内的平面感知与平面间深度测量的能力。通过与现有双目散斑结构光法及商用双目散斑深度相机的对比试验证明,文中方法能在较大深度范围内稳定完成立体匹配,具备较高的精度和鲁棒性,在工业三维感知领域存在良好的发展潜力。 展开更多
关键词 结构光 双目视觉 半全局立体匹配 三维重建 图像处理
在线阅读 下载PDF
上一页 1 2 250 下一页 到第
使用帮助 返回顶部