期刊文献+
共找到5,878篇文章
< 1 2 250 >
每页显示 20 50 100
动态场景下基于跨域掩膜分割的视觉SLAM算法
1
作者 亢洁 徐婷 +4 位作者 王佳乐 郭进 赫轩 王沫 夏宇 《陕西科技大学学报》 北大核心 2026年第1期178-185,193,共9页
针对动态场景下视觉SLAM(Simultaneous Localization and Mapping)系统中深度学习分割网络实时性不足,以及相机非期望运动导致位姿估计偏差的问题,提出一种基于跨域掩膜分割的视觉SLAM算法.该算法采用轻量化YOLO-fastest网络结合背景减... 针对动态场景下视觉SLAM(Simultaneous Localization and Mapping)系统中深度学习分割网络实时性不足,以及相机非期望运动导致位姿估计偏差的问题,提出一种基于跨域掩膜分割的视觉SLAM算法.该算法采用轻量化YOLO-fastest网络结合背景减除法实现运动物体检测,利用深度图结合深度阈值分割构建跨域掩膜分割机制,并设计相机运动几何校正策略补偿检测框坐标误差,在实现运动物体分割的同时提升处理速度.为优化特征点利用率,采用金字塔光流对动态特征点进行帧间连续跟踪与更新,同时确保仅由静态特征点参与位姿估计过程.在TUM数据集上进行系统性评估,实验结果表明,相比于ORB-SLAM3算法,该算法的绝对位姿误差平均降幅达97.1%,与使用深度学习分割网络的DynaSLAM和DS-SLAM的动态SLAM算法相比,其单帧跟踪时间大幅减少,在精度与效率之间实现了更好的平衡. 展开更多
关键词 视觉SLAM 动态场景 YOLO-Fastest 金字塔光流 深度阈值分割
在线阅读 下载PDF
基于多注意力机制的脊柱病灶MRI影像识别模型
2
作者 周慧 宋新景 《计算机科学与探索》 北大核心 2026年第1期291-300,共10页
人工检测脊柱病变是一项耗时的工作,并且高度依赖于该领域的专家,因此脊柱病灶的自动识别是非常必要的。然而,因为脊柱病灶的大小、位置和结构存在着广泛的差异,同时脊柱肿瘤与稀有病布鲁氏菌在影像上高度相似,所以脊柱病灶的准确定位... 人工检测脊柱病变是一项耗时的工作,并且高度依赖于该领域的专家,因此脊柱病灶的自动识别是非常必要的。然而,因为脊柱病灶的大小、位置和结构存在着广泛的差异,同时脊柱肿瘤与稀有病布鲁氏菌在影像上高度相似,所以脊柱病灶的准确定位和分类是一项具有挑战性的工作。为了应对这些挑战,提出了一种改进的脊柱病灶MRI影像识别模型。引入以ResNet-101为基础的双向特征金字塔主干网络,利用可变卷积在不同层替代传统的卷积神经网络,从特征层中获得更多的特征信息。在不同的模块中加入了多重注意力,包括自注意力机制和柔性注意力机制,有效地融合特征中贡献较大的部分。为了克服脊柱肿瘤、感染性病变、稀有病布鲁氏菌的数据不平衡问题,引入了改进的平衡交叉熵损失函数。在大连某医院提供的临床数据集上进行验证,识别精确率达到了94.2%,识别召回率达到90.8%。与其他识别模型进行对比实验,结果说明了该方法相对于其他模型识别性能更好。 展开更多
关键词 脊柱病灶识别 双向特征金字塔 多注意力机制 可变卷积 多特征融合
在线阅读 下载PDF
Enhancement of Rice Germplasm by Pyramiding the Cultivars with Multi-resistance to Orseolia oryzae (Wood-Mason) 被引量:2
3
作者 冯锐 秦学毅 +5 位作者 唐建淮 朱汝财 潘英华 刘百龙 韦素美 黄凤宽 《Agricultural Science & Technology》 CAS 2009年第1期7-10,共4页
[ Objective] This study was to breed rice cultivars with multi-resistance to Orseolia oryzae (Wood-Mason). [ Method] The Guangxi local cultivar GX-M001 (Jiangchao) with high resistance to Orseolia oryzae (Wood-Ma... [ Objective] This study was to breed rice cultivars with multi-resistance to Orseolia oryzae (Wood-Mason). [ Method] The Guangxi local cultivar GX-M001 (Jiangchao) with high resistance to Orseolia oryzae (Wood-Mason) was used to hybrid with the known resistance cultivars "Kangwenqingzhan" (harboring GM5 gene), OB677( harboring GM3 gene) from Sri Lanka, HT1350 and high yield end quality cultivar " Guiruanzhan". [ Result] Through pyramiding the multi-resistant genes via routine hybridization, the general resistances of the hybrids were remarkably enhanced. The grades of resistance were also improved, many of the combinations were endowed with a resistance at immune level (grade 0) ; and interestingly, the respective hybridization of GX-M001 (high resistance) with OB677( medium resistance) and HT1350(suscepti- ble) also generate two lines at immune level, which is probably the effects of additive effects of genes.[ Conclusion] By routine hybridization, multiple genes were successfully pyramided, thus generating novel rice lines with multiple resistances. For the rice breeding scientists at the grass-roots level, the resistance-resistance pyramiding is an effective approach to breed high resistance cultivars. 展开更多
关键词 Orseolia oryzae (Wood-Mason) pyramiding via hybridization Rice breeding
在线阅读 下载PDF
基于空间通道自适应特征的肝脏病理图像分割网络
4
作者 王建宇 王朝立 +1 位作者 孙占全 刘晓虹 《电子科技》 2026年第1期9-17,共9页
针对肝脏病理图像中病变区域与周围组织相似度高、对比度低以及边界模糊等问题,文中提出了一个基于空间通道自适应特征的肝脏病理分割网络。通过混合校准注意力使网络能够自适应地选择经空间和通道校准过的特征信息,有利于编码器捕获与... 针对肝脏病理图像中病变区域与周围组织相似度高、对比度低以及边界模糊等问题,文中提出了一个基于空间通道自适应特征的肝脏病理分割网络。通过混合校准注意力使网络能够自适应地选择经空间和通道校准过的特征信息,有利于编码器捕获与肝脏病灶相关的重要特征,并在编码器最深层引入空洞空间金字塔池化模块来弥补高级特征所缺失的多尺度信息,提高模型的分割精度。在私有肝脏数据集、公开肝脏数据集以及其他两种公开病理数据集对所提网络进行对比实验和消融实验。实验结果表明,相较于其他方法,所提网络的分割结果较佳,且有效解决了肝细胞癌分割问题。 展开更多
关键词 肝细胞癌 病理图像 编解码架构 混合校准注意力模块 空间注意力 通道注意力 空洞空间金字塔池化模块 多尺度信息
在线阅读 下载PDF
基于改进Faster R—CNN的水稻秧苗漏插识别研究
5
作者 邹立雯 梁春英 +2 位作者 胡军 陈玉恒 李圳鹏 《中国农机化学报》 北大核心 2026年第2期101-107,共7页
水稻是我国的主要粮食作物,实现水稻的高产优产是必然趋势。针对传统人工补苗效率低、主观性高的问题,提出一种基于改进Faster R—CNN的水稻秧苗漏插识别方法。以Faster R—CNN模型为基础,将主干网络替换为残差网络ResNet50,结合FPN特... 水稻是我国的主要粮食作物,实现水稻的高产优产是必然趋势。针对传统人工补苗效率低、主观性高的问题,提出一种基于改进Faster R—CNN的水稻秧苗漏插识别方法。以Faster R—CNN模型为基础,将主干网络替换为残差网络ResNet50,结合FPN特征金字塔对特征信息进行提取;引入RoI Align双线性插值的思想替代RoI Pooling层粗糙量化操作。结果表明,改进后的Faster R—CNN模型识别的精确率为93.62%,平均精度均值mAP@0.5为95.06%;与未改进的模型相比,识别精确率提高7.33%,模型的平均精度均值mAP@0.5提高4.6%。该模型可以提高水稻秧苗的分类和插秧机漏插位置的检测精度,为制定水稻秧苗补苗计划打下坚实的基础,并为评价水稻插秧机质量提供数据支持。 展开更多
关键词 水稻秧苗 漏插识别 特征金字塔 深度学习 残差网络
在线阅读 下载PDF
EHDC-YOLO: Enhancing Object Detection for UAV Imagery via Multi-Scale Edge and Detail Capture
6
作者 Zhiyong Deng Yanchen Ye Jiangling Guo 《Computers, Materials & Continua》 2026年第1期1665-1682,共18页
With the rapid expansion of drone applications,accurate detection of objects in aerial imagery has become crucial for intelligent transportation,urban management,and emergency rescue missions.However,existing methods ... With the rapid expansion of drone applications,accurate detection of objects in aerial imagery has become crucial for intelligent transportation,urban management,and emergency rescue missions.However,existing methods face numerous challenges in practical deployment,including scale variation handling,feature degradation,and complex backgrounds.To address these issues,we propose Edge-enhanced and Detail-Capturing You Only Look Once(EHDC-YOLO),a novel framework for object detection in Unmanned Aerial Vehicle(UAV)imagery.Based on the You Only Look Once version 11 nano(YOLOv11n)baseline,EHDC-YOLO systematically introduces several architectural enhancements:(1)a Multi-Scale Edge Enhancement(MSEE)module that leverages multi-scale pooling and edge information to enhance boundary feature extraction;(2)an Enhanced Feature Pyramid Network(EFPN)that integrates P2-level features with Cross Stage Partial(CSP)structures and OmniKernel convolutions for better fine-grained representation;and(3)Dynamic Head(DyHead)with multi-dimensional attention mechanisms for enhanced cross-scale modeling and perspective adaptability.Comprehensive experiments on the Vision meets Drones for Detection(VisDrone-DET)2019 dataset demonstrate that EHDC-YOLO achieves significant improvements,increasing mean Average Precision(mAP)@0.5 from 33.2%to 46.1%(an absolute improvement of 12.9 percentage points)and mAP@0.5:0.95 from 19.5%to 28.0%(an absolute improvement of 8.5 percentage points)compared with the YOLOv11n baseline,while maintaining a reasonable parameter count(2.81 M vs the baseline’s 2.58 M).Further ablation studies confirm the effectiveness of each proposed component,while visualization results highlight EHDC-YOLO’s superior performance in detecting objects and handling occlusions in complex drone scenarios. 展开更多
关键词 UAV imagery object detection multi-scale feature fusion edge enhancement detail preservation YOLO feature pyramid network attention mechanism
在线阅读 下载PDF
改进YOLOv8_obb的大豆主茎节点识别研究
7
作者 杨彦旭 李金阳 +2 位作者 石文强 亓立强 张伟 《中国农机化学报》 北大核心 2026年第1期79-86,共8页
大豆株型对大豆产量有重要影响,大豆主茎节数是大豆株型构成的重要性状。为实现田间条件下大豆主茎节数识别计算,以黑龙江省九三地区大豆为研究对象,提出基于YOLOv8_obb模型改进的大豆主茎节点识别方法YOLOv8_obb—AES,计算大豆主茎节点... 大豆株型对大豆产量有重要影响,大豆主茎节数是大豆株型构成的重要性状。为实现田间条件下大豆主茎节数识别计算,以黑龙江省九三地区大豆为研究对象,提出基于YOLOv8_obb模型改进的大豆主茎节点识别方法YOLOv8_obb—AES,计算大豆主茎节点,得到大豆主茎节数。改进模型引入高效注意力机制模块,缩减模型计算量,采用渐进特征金字塔网络结构替换YOLOv8_obb网络中的路径聚合特征金字塔网络,增强多尺度融合能力,替换IoU损失函数加快模型边界回归,提高模型收敛速度。结果表明,YOLOv8_obb—AES算法对田间大豆主茎节点的平均精度均值与检测速度分别达到89.45%、78.8帧/ms,相比于原始算法分别提升8.45%、7.6帧/ms,对于九研17大豆植株6种不同主茎节数的识别准确率分别为85.4%、84.5%、87.6%、85.2%、81.6%和82.2%。该研究为探究大豆产量与大豆主茎节数之间的关联提供技术支持。 展开更多
关键词 大豆 主茎节点 目标识别 渐进特征金字塔网络 高效注意力机制
在线阅读 下载PDF
基于跨尺度特征融合的内窥镜图像增强算法
8
作者 刘旭阳 蔡芸 蒋林 《现代电子技术》 北大核心 2026年第1期34-40,共7页
临床医学的内窥镜图像由于在成像过程中存在补充光源不均匀和人体组织粘液反光的问题,出现大量曝光过度等图像质量较低的现象。现有基于深度学习的图像增强算法由于仅采用固定尺寸的特征融合方式,导致特征提取能力较低、增强效果较差。... 临床医学的内窥镜图像由于在成像过程中存在补充光源不均匀和人体组织粘液反光的问题,出现大量曝光过度等图像质量较低的现象。现有基于深度学习的图像增强算法由于仅采用固定尺寸的特征融合方式,导致特征提取能力较低、增强效果较差。为改善这一问题,文中构建了基于跨尺度特征融合的内窥镜图像增强算法,通过构建CM卷积模块实现高性能特征提取,同时采用SPPF金字塔池化模块实现对特征图不同尺度的池化操作,并且在网络不同尺度的网络层之间引入跨尺度特征融合(CFF)模块,实现多尺度特征融合和上下文信息传播,从而大幅提高图像细节捕捉能力和图像质量。实验结果表明,文中算法在PSNR、SSIM指标均高于现有算法,其中PSNR指标提高了9.9%,SSIM指标提高了15.4%,可以实现高质量内窥镜图像增强任务。 展开更多
关键词 内窥镜图像 深度特征融合 CFF 曝光异常 图像增强算法 金字塔池化模块
在线阅读 下载PDF
FireLight-YOLO:面向森林火灾实时监测的轻量化模型
9
作者 李敏学 张晓宇 +2 位作者 程英杰 霍光煜 许福 《北京林业大学学报》 北大核心 2026年第1期12-25,共14页
【目的】为应对森林火灾频发对生态安全构成的严峻挑战,构建轻量化实时智能监测体系以提升生态风险防控能力具有重要现实意义。针对现有火灾检测方法易受环境干扰,且模型复杂度与实时性难以兼顾的问题,本研究旨在开发一种无需外部预训... 【目的】为应对森林火灾频发对生态安全构成的严峻挑战,构建轻量化实时智能监测体系以提升生态风险防控能力具有重要现实意义。针对现有火灾检测方法易受环境干扰,且模型复杂度与实时性难以兼顾的问题,本研究旨在开发一种无需外部预训练权重即可从零训练的高效轻量化检测模型。【方法】研究首先构建了涵盖1万余张高质量图像的森林火灾监测数据集并开源发布。在此基础上,基于YOLOv8提出FireLight-YOLO轻量化架构:引入幽灵卷积压缩冗余计算,设计融合部分卷积与点态卷积的FasterC2fBlock构建T形感受野以增强关键区域感知,并优化SPPF模块提出特征金字塔共享卷积机制实现高效跨尺度特征融合。模型通过交叉验证、独立测试、消融实验及多噪声场景鲁棒性检验完成性能评估。【结果】FireLight-YOLO在未使用预训练权重条件下实现mAP@0.5达0.491,仅需约2.26×10^(6)参数与5.9GFLOPs计算量,在精度、轻量化与实时性间达到有效平衡。相较于原始YOLOv8,模型计算量减少2.2 GFLOPs,参数量降低了25%,推理速度提升15%,并在复杂干扰场景中展现出优异的鲁棒性。【结论】FireLight-YOLO实现了轻量化条件下对森林火灾的精准检测。该研究不仅为森林火灾智能监测提供了低成本、高效率的技术方案,其轻量化特性亦显著增强了模型在移动终端的部署适应性。研究成果可为森林生态系统的保护与修复提供坚实的智能化支撑。 展开更多
关键词 YOLOv8 Ghost卷积 森林火灾检测 实时目标检测 轻量化模型 特征金字塔共享卷积(FPSC) 边缘部署 生态安全
在线阅读 下载PDF
DEPA-YOLO:无人机视角下的小目标检测模型
10
作者 刘臣杰 刘巍 +1 位作者 杨雯迪 王成 《计算机科学与探索》 北大核心 2026年第1期266-279,共14页
针对无人机视角下小目标检测中存在的目标密集分布、背景复杂干扰及特征分辨率不足等问题,提出了一种基于YOLOv10的改进模型DEPA-YOLO。该模型设计了动态多尺度特征提取模块DCMB,该模块通过融合动态权重分配与多形态特征混合策略,有效... 针对无人机视角下小目标检测中存在的目标密集分布、背景复杂干扰及特征分辨率不足等问题,提出了一种基于YOLOv10的改进模型DEPA-YOLO。该模型设计了动态多尺度特征提取模块DCMB,该模块通过融合动态权重分配与多形态特征混合策略,有效增强了浅层局部纹理与高层全局语义的联合建模能力,从而改善小目标的特征表达。提出了HFEP特征金字塔结构,将SPDConv与ECFI的多分支动态融合相结合,实现了对不同尺度目标的细节保持与多层次语义的高效传递,显著缓解了传统特征金字塔在密集场景中信息丢失的问题;在主干网络中引入CPCA注意力机制,引导网络聚焦于显著目标区域,减少复杂背景干扰带来的误检与漏检;并在损失函数中采用WIoU回归策略,通过引入梯度调控与边界惩罚机制,提高了小目标定位的稳定性与鲁棒性。在VisDrone2019数据集上的消融实验结果表明,DEPA-YOLO相较YOLOv10基准模型在精度、mAP50和mAP50:95上分别提升4.7、4.3和2.5个百分点;在DOTA数据集的泛化实验中进一步提升7.9、2.0和1.2个百分点。DEPA-YOLO模型显著提升了无人机视角下的小目标检测精度,为无人机小目标检测提供了一种高效的解决方案。 展开更多
关键词 小目标检测 YOLOv10 特征增强金字塔 注意力机制
在线阅读 下载PDF
Pyramiding blast,bacterial blight and brown planthopper resistance genes in rice restorer lines 被引量:6
11
作者 JI Zhi-juan Yang Shu-dong +3 位作者 ZENG Yu-xiang LIANG Yan YANG Chang-deng QIAN Qian 《Journal of Integrative Agriculture》 SCIE CAS CSCD 2016年第7期1432-1440,共9页
Rice blast, bacterial blight (BB) and brown planthopper (BPH) are the three main pests of rice. This study investigated pyr-amiding genes resistant to blast, BB and BPH to develop restorer lines. Ten new lines wit... Rice blast, bacterial blight (BB) and brown planthopper (BPH) are the three main pests of rice. This study investigated pyr-amiding genes resistant to blast, BB and BPH to develop restorer lines. Ten new lines with blast, BB and/or BPH resistance genes were developed using marker-assisted selection (MAS) technique and agronomic trait selection (ATS) method. Only HR13 with resistance genes to blast, BB and BPH was obtained. In addition to blast and BB resistance, four lines (HR39, HR41, HR42, HR43) demonstrated moderate resistance to BPH, but MAS for BPH resistance genes were not conducted in developing these four lines. These data suggested that there were unknown elite BPH resistance genes in the Zhongzu 14 donor parent. A more effective defense was demonstrated in the lines withPi1 andPi2 genes although the weather in 2012 was favorable to disease incidence. Blast resistance of the lines with a single resistance gene,Pita, was easily inlfuenced by the weather. Overal, the information obtained through pyramiding multiple resistance genes on developing the restorer lines is helpful for rice resistance breeding. 展开更多
关键词 RICE BLAST bacterial blight brown planthopper resistance PYRAMID
在线阅读 下载PDF
Marker-Assisted Pyramiding of Genes Conferring ResistanceAgainst Bacterial Blight and Blast Diseases into Indian RiceVariety MTU1010 被引量:6
12
作者 k.arunakumari c.v.durgarani +10 位作者 v.satturu k.r.sarikonda p.d.r.chittoor b.vutukuri g.s.laha a.p.k.nelli s.gattu m.jamal a.prasadbabu s.hajira r.m.sundaram 《Rice science》 SCIE CSCD 2016年第6期306-316,共11页
Two major bacterial blight (BB) resistance genes (Xa21 and xa13) and a major gene for blastresistance (Pi54) were introgressed into an Indian rice variety MTU1010 through marker-assistedbackcross breeding. Impro... Two major bacterial blight (BB) resistance genes (Xa21 and xa13) and a major gene for blastresistance (Pi54) were introgressed into an Indian rice variety MTU1010 through marker-assistedbackcross breeding. Improved Samba Mahsuri (possessing Xa21 and xa13) and NLR145 (possessingPi54) were used as donor parents. Marker-assisted backcrossing was continued till BC2 generationwherein PCR based functional markers specific for the resistance genes were used for foregroundselection and a set of parental polymorphic microsatellite markers were used for background selectionat each stage of backcrossing. Selected BC2F1 plants from both crosses, having the highest recoveriesof MTU1010 genome (90% and 92%, respectively), were intercrossed to obtain intercross F1 (ICF1) plants,which were then selfed to generate 880 ICF2 plants possessing different combinations of the BB andblast resistance genes. Among the ICF2 plants, seven triple homozygous plants (xa13xa13Xa21Xa21Pi54Pi54)with recurrent parent genome recovery ranging from 82% to 92% were identified. All the seven ICF2plants showed high resistance against the bacterial blight disease with a lesion lengths of only 0.53–2.28 cm, 1%–5% disease leaf areas and disease scoring values of ‘1’ or ‘3’. The seven ICF2 plants wereselfed to generate ICF3, which were then screened for blast resistance, and all were observed to behighly resistant to the diseases. Several ICF3 lines possessing high level of resistance against BB andblast, coupled with yield, grain quality and plant type on par with MTU1010 were identified and advanced forfurther selection and evaluation. 展开更多
关键词 gene pyramiding bacterial blight resistance blast resistance RICE marker-assistedbackcross breeding
在线阅读 下载PDF
Marker-assisted pyramiding of soybean resistance genes R_(SC4),R_(SC8),and R_(SC14Q) to soybean mosaic virus 被引量:5
13
作者 WANG Da-gang ZHAO Lin +5 位作者 LI Kai MA Ying WANG Li-qun YANG Yong-qing YANG Yun-hua ZHI Hai-jian 《Journal of Integrative Agriculture》 SCIE CAS CSCD 2017年第11期2413-2420,共8页
Soybean mosaic virus (SMV) is one of the major viral pathogens affecting soybean crops worldwide. Three SMV resistance genes, Rsc4, Rsc8, and Rsc14Q, have been identified and mapped on soybean chromosomes 14, 2, and... Soybean mosaic virus (SMV) is one of the major viral pathogens affecting soybean crops worldwide. Three SMV resistance genes, Rsc4, Rsc8, and Rsc14Q, have been identified and mapped on soybean chromosomes 14, 2, and 13 from Dabaima, Kefeng 1, and Qihuang 1 cultivars, respectively. Soybean cultivar Nannong 1138-2 is widely grown in the Yangtze River Valley of China. In this study, crosses were made between Qihuang l^Kefeng 1 and DabaimaxNannong 1138-2. Ten simple sequence repeat (SSR) markers linked to three resistance loci (Rsc4, Rsc8, and Rsc^4Q) were used to assist pyramided breeding. Pyramided families containing three resistance loci (Rsc4, Rsc8, and Rsc14Q) were evaluated by inoculating them with 21 SMV strains from China. Results indicated that the 10 markers can be used effectively to assist the selection of resistant individuals containing Rsc4, Rsc8, and Rsc14Q. A total of 53 F6 plants were confirmed to contain three homozygous alleles conferring resistance to SMV. Five F7 homozygous pyramided families exhibited resistance to 21 strains of SMV and showed desirable agronomic traits using dual selection. The strategy of pyramiding resistance gene derived from different varieties has practical breeding value in providing broad-spectrum resistance against the existing strains of SMV in China. 展开更多
关键词 SOYBEAN soybean mosaic virus resistance genes pyramiding marker-assisted breeding
在线阅读 下载PDF
Genetic Improvement of Japonica Rice Variety Wuyujing 3 for Stripe Disease Resistance and Eating Quality by Pyramiding Stv-b^i and Wx-mq 被引量:5
14
作者 CHEN Tao WU Hao +8 位作者 ZHANG Ya-dong ZHU Zhen ZHAO Qi-yong ZHOU Li-hui YAO Shu ZHAO Ling YU Xin ZHAO Chun-fang WANG Cai-lin 《Rice science》 SCIE CSCD 2016年第2期69-77,共9页
Japonica rice variety Kanto 194 as the donor of resistance gene Stv-b^i and Iow-amylose content gene Wx-mq was used to improve the resistance and eating quality of Wuyujing 3 by the breeding strategy of backcross. In ... Japonica rice variety Kanto 194 as the donor of resistance gene Stv-b^i and Iow-amylose content gene Wx-mq was used to improve the resistance and eating quality of Wuyujing 3 by the breeding strategy of backcross. In continuous backcross and selfcross generations, the related molecular markers with Stvobi and Wx-mq genes were utilized for genotypic detection by associated with resistance identification of rice stripe disease and agronomic traits selection. Finally, 10 improved lines with homozygous genotype Stv-b^i Stv-b^iWx-mqWx-mq were obtained from BC3F4 generation. The results of comparative analysis indicated that most characters of these lines were consistent with those of recipient parent Wuyujing 3, except for the improved resistance, appearance and eating quality. By evaluation of the comprehensive performance of them, two excellent lines K01 and K04 were selected for further experiments. 展开更多
关键词 rice stripe disease resistance eating quality marker-assisted pyramiding selection genetic improvement
在线阅读 下载PDF
Simulation of gene pyramiding in Drosophila melanogaster 被引量:4
15
作者 Li Jiang Fuping Zhao Qin Zhang 《Journal of Genetics and Genomics》 SCIE CAS CSCD 北大核心 2008年第12期737-742,共6页
Gene pyramiding has been successfully practiced in plant breeding for developing new breeds or lines in which favorable genes from several different lines were integrated. But it has not been used in animal breeding, ... Gene pyramiding has been successfully practiced in plant breeding for developing new breeds or lines in which favorable genes from several different lines were integrated. But it has not been used in animal breeding, and some theoretical investigation and simulation analysis with respect to its strategies, feasibility and efficiency are needed before it can be implemented in animals. In this study, we used four different pure lines of Drosophila melanogaster, each of which is homozygous at a specific mutant gene with a visible effect on phenotype, to simulate the gene pyramiding process and analyze the duration and population size required in different pyramiding strategies. We finally got the ideal individuals, which are homozygous at the four target genes simultaneously. This study demonstrates that gene pyramiding is feasible in animal breeding and the interaction between genes may affect the final results. 展开更多
关键词 gene pyramiding Drosophila melanogaster SIMULATION
在线阅读 下载PDF
Pyramiding of multiple genes generates rapeseed introgression lines with clubroot and herbicide resistance, high oleic acid content, and early maturity 被引量:4
16
作者 Zhaoyang Wang Fucai Wang +6 位作者 Zihan Yu Xiaorui Shi Xianming Zhou Pengfei Wang Yixian Song Dengfeng Hong Guangsheng Yang 《The Crop Journal》 SCIE CSCD 2023年第3期895-903,共9页
Clubroot and herbicide resistance,high oleic acid(OA)content,and early maturity are targets of rapeseed(Brassica napus L.)breeding.The objective of this study was to develop new male-fertility restorer lines by pyrami... Clubroot and herbicide resistance,high oleic acid(OA)content,and early maturity are targets of rapeseed(Brassica napus L.)breeding.The objective of this study was to develop new male-fertility restorer lines by pyramiding favorable genes to improve these traits simultaneously.Seven elite alleles for the four traits were introduced into the restorer line 621R by speed breeding with marker-assisted and phenotypic selection.Six introgression lines(ILs)were developed with four-to seven-gene combinations and crossed with two elite parents to develop hybrids.All ILs and their corresponding hybrids displayed high resistance to both clubroot pathotype 4 and sulfonylurea herbicides.Three ILs and their hybrids showed large increases in OA contents and four showed earlier maturity.These new ILs may be useful in rapeseed hybrid breeding for the target traits. 展开更多
关键词 Clubroot resistance Herbicide resistance Oleic acid content Flowering time Pyramid breeding
在线阅读 下载PDF
Application of Molecular Marker Assisted Selection in Gene Pyramiding and Selection of New Cultivars 被引量:4
17
作者 WU Lan WANG Chao 《Journal of Northeast Agricultural University(English Edition)》 CAS 2011年第1期79-84,共6页
The feasibility of molecule markers' application in gene pyramiding has been proved,and obvious progresses in crop breeding have been made till now.Furthermore,different QTLs or molecular markers linked tightly to yi... The feasibility of molecule markers' application in gene pyramiding has been proved,and obvious progresses in crop breeding have been made till now.Furthermore,different QTLs or molecular markers linked tightly to yield,quality or resistance may be used for marker assisted selection.MAS will be applied widely in crop breeding due to the development of more gene-based markers and efficient quantitative trait locus(QTL) as well as lower cost marking systems. 展开更多
关键词 molecular marker gene pyramiding BREEDING
在线阅读 下载PDF
融合金字塔结构与注意力机制的煤矿井下作业安全监控方法
18
作者 任延川 《煤矿现代化》 2026年第1期149-153,共5页
煤矿井下作业安全监控过程中,主要依靠固定传感检测,监测数据传输困难,导致作业异常安全监控预警频率下降,为此提出对融合金字塔结构与注意力机制的煤矿井下作业安全监控方法的设计分析。在煤矿井下建立局部坐标系,部署监控节点,通过坐... 煤矿井下作业安全监控过程中,主要依靠固定传感检测,监测数据传输困难,导致作业异常安全监控预警频率下降,为此提出对融合金字塔结构与注意力机制的煤矿井下作业安全监控方法的设计分析。在煤矿井下建立局部坐标系,部署监控节点,通过坐标系和局部监控节点确定可监控空间,在该空间之内,将局部坐标系转换为全局坐标系,融合金字塔结构,将全局坐标系中节点采集的图像归类,作为可应用条件。利用采集的图像,进行空间特征的提取,明确空间注意力后,以特征为引导,映射关键监控区域,在该区域内计算出全局坐标系中各点位之间的安全监控均方误差。增设动态预警规则,根据监控均方误差的变化,利用规则判定监控情况是否存在异常,输出最终的监控结果。实验结果表明:所提方法得出的作业异常安全监控预警频率均可以达到205 Hz以上,说明对作业实况异常的感知速度较快,优越性较强。 展开更多
关键词 金字塔结构 注意力机制 井下作业 安全监控 煤矿定位
在线阅读 下载PDF
基于改进YOLOv12的无人机高速公路边坡起火检测算法研究
19
作者 崔培强 罗梦琪 +3 位作者 刘世达 欧阳奕波 郭建 陈少坤 《电子技术应用》 2026年第1期14-20,共7页
针对无人机高速公路边坡火灾检测中存在的初期火苗漏检率高、复杂背景干扰显著以及边缘设备部署精度-实时性难以平衡等关键问题,基于改进YOLOv12框架提出一种轻量化可见光火灾检测算法。该算法通过实例分割技术精准提取边坡与道路区域... 针对无人机高速公路边坡火灾检测中存在的初期火苗漏检率高、复杂背景干扰显著以及边缘设备部署精度-实时性难以平衡等关键问题,基于改进YOLOv12框架提出一种轻量化可见光火灾检测算法。该算法通过实例分割技术精准提取边坡与道路区域以消除背景干扰,设计包含高分辨率P2层与多路径下采样机制的小目标增强特征金字塔(SOPAN)提升微小目标检测能力,并构建基于共享卷积的轻量检测头(LSCD)实现模型复杂度压缩。实验结果表明,在融合无人机巡检数据集上,所提模型以仅1.8M参数量达到92.5%mAP检测精度,较基准模型YOLOv12n参数量降低28.3%,为无人机端实时高精度火灾预警提供了高效解决方案。 展开更多
关键词 YOLOv12 小目标增强特征金字塔 轻量化检测 边坡起火检测
在线阅读 下载PDF
Stimulation Study of Gene Pyramiding inAnimals by Marker-Assisted Selection 被引量:2
20
作者 ZHAO Fu-ping ZHANG Qin 《Journal of Integrative Agriculture》 SCIE CAS CSCD 2012年第11期1871-1876,共6页
This gene pyramiding strategy is based on the idea of efficiently pyramiding genes of interest by crosses and selection to obtain a population with favorable alleles from different breeds or lines, which is called an ... This gene pyramiding strategy is based on the idea of efficiently pyramiding genes of interest by crosses and selection to obtain a population with favorable alleles from different breeds or lines, which is called an ideal population. We investigate impacts of some factors on the pyramiding efficiencies by simulation. These factors include selection strategies (the breeding value selection, the molecular scores selection and the index selection), proportion selected (2, 10 and 20%), recombination rates between adjacent target genes (0.1, 0.3 and 0.5) and different mating types (the random mating and the positive assortative mating avoiding sib mating). The results show that: (1) The more recombination rate and the lower proportion male selected, the better pyramiding efficiency; (2) the ideal population is obtained via various selection strategies, while different selection strategies are suitable for different breeding objectives. From the perspective of pyramiding target genes merely, the molecular scores selection is the best one, for the purpose of pyramiding target genes and recovering genetic background of the target trait, the index selection is the best one, while from the saving cost point of view, the breeding value selection is the best one; (3) the positive assortative mating is more efficient for gene pyramiding compared with the random mating in the terms of the number of generations of intercross for getting the ideal population. 展开更多
关键词 gene pyramiding pyramiding efficiency selection strategies mating types
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