A fast algorithm based on the grayscale distribution of infrared target and the weighted kernel function was proposed for the moving target detection(MTD) in dynamic scene of image series. This algorithm is used to de...A fast algorithm based on the grayscale distribution of infrared target and the weighted kernel function was proposed for the moving target detection(MTD) in dynamic scene of image series. This algorithm is used to deal with issues like the large computational complexity, the fluctuation of grayscale, and the noise in infrared images. Four characteristic points were selected by analyzing the grayscale distribution in infrared image, of which the series was quickly matched with an affine transformation model. The image was then divided into 32×32 squares and the gray-weighted kernel(GWK) for each square was calculated. At last, the MTD was carried out according to the variation of the four GWKs. The results indicate that the MTD can be achieved in real time using the algorithm with the fluctuations of grayscale and noise can be effectively suppressed. The detection probability is greater than 90% with the false alarm rate lower than 5% when the calculation time is less than 40 ms.展开更多
We prove some value-distribution results for a class of L-functions with rational moving targets. The class contains Selberg class, as well as the Riemann-zeta function.
Parkinson＇s disease （PD） is a progressive neurodegenerative disease, which is generally considered a multifactorial disorder that arises owing to a combination of genes and environmental factors. While most cases a...Parkinson＇s disease （PD） is a progressive neurodegenerative disease, which is generally considered a multifactorial disorder that arises owing to a combination of genes and environmental factors. While most cases are idiopathic, in about 10% of the patients a genetic cause can be detected, ascribable to mutations in more than a dozen genes. PD is characterized clinically by tremor, rigidity, reduced mo- tor activity （bradykinesia）, and postural instability and pathological- ly by loss of dopaminergic （DA） neurons in the substantia nigra pars compacta, loss of DA innervation in the striatum, and the presence of a-synuclein positive aggregates in the form of Lewy bodies. The symptomatic treatment of PD with levodopa, which aims at replac- ing dopamine, remains the gold standard, and no neuroprotective or disease-modifying therapy is available. During treatment, the disease continues to progress, and long-term use of levodopa has import- ant limitations including motor complications termed dyskinesias. Therefore, a pharmacological therapy able to prevent or halt the neu- rodegenerative process is urgently required.展开更多
In this paper, a novel signal processing technique hasbeen developed to refocus moving targets image from their smeared responses in the Synthetic Aperture Radar (SAR) image according to the characteristics of the rec...In this paper, a novel signal processing technique hasbeen developed to refocus moving targets image from their smeared responses in the Synthetic Aperture Radar (SAR) image according to the characteristics of the received signals for moving targets. Quadratic Phase Function is introduced to the parameters estimation for moving target echo and SAR imaging. Our method is available even under a low SNR environment and acquiring an exact SAR image of moving targets. The simulated results demonstrated the validity of the algorithm proposed.展开更多
针对朝天椒检测深度学习网络模型体积大、参数多以及在计算资源有限的移动设备中难以部署等问题,文中提出一种基于YOLOv5s(You Only Look Once version 5s)的轻量级朝天椒检测模型。利用GhostNet中Ghost模块和Ghost瓶颈(Ghost BottleNe...针对朝天椒检测深度学习网络模型体积大、参数多以及在计算资源有限的移动设备中难以部署等问题,文中提出一种基于YOLOv5s(You Only Look Once version 5s)的轻量级朝天椒检测模型。利用GhostNet中Ghost模块和Ghost瓶颈(Ghost BottleNeck)结构重新构建特征提取网络,减少模型的参数量和计算复杂度。在特征融合部分采用GSConv(Ghost-Shuffle Convolution)轻量级卷积和VoV-GSCSP(VoVNet-Ghost Shuffle-Cross Stage Partial)结构分别替换原始卷积和CSP(Cross Stage Partial)模块,在保证精度的同时使模型轻量化效果最佳。采用角度惩罚度量的SIoU(SCYLLA-Intersaction over Union)损失优化边界框损失函数,提升了轻量化后的模型精度和泛化能力。实验结果表明,改进YOLOv5s-GGS(YOLOv5s-GhostNet GSConv SIoU)模型相较于原始网络模型的精确度、召回率和平均精度均值(mean Average Precision,mAP)分别提高了7.0百分点、3.5百分点和3.8百分点,参数量、计算复杂度和权重降低了42%以上。相较于主流目标检测模型,所提模型具有更高的检测精度以及更少的模型体积,实现了模型的轻量化,精度提升较大,推理速度较快,更适合部署于移动设备。展开更多
切段式机收甘蔗含杂率的自动测量可以客观评估机收甘蔗到糖厂入榨前的质量。针对现有抽样称重估算杂质方式效率低且主观性强的问题,以及因田间环境较为复杂使得检测目标蔗段存在运动状态变换导致的模糊、光照强度变化和蔗叶遮挡等技术难...切段式机收甘蔗含杂率的自动测量可以客观评估机收甘蔗到糖厂入榨前的质量。针对现有抽样称重估算杂质方式效率低且主观性强的问题,以及因田间环境较为复杂使得检测目标蔗段存在运动状态变换导致的模糊、光照强度变化和蔗叶遮挡等技术难点,提出了一种基于改进YOLOv5安装在切段式甘蔗机上的机收蔗含杂率检测的方法。首先,针对工业相机拍摄的蔗段目标为小目标的应用场景,增加小目标检测层,增强网络模型对其的专注;其次,将C3模块替换成C2f模块,提高网络模型对小物体、低对比度目标的检测速度和检测精度;最后,加入加权交并比WIoU(Weighted Intersection over Union)损失函数,提升预测框的回归精度,增强数据集训练效果。试验结果表明:基于改进YOLOv5的机收蔗含杂率检测模型,蔗段识别准确率达95.2%、mAP(mean Average Precision)值为62.5%,相较于原始YOLOv5模型分别提高了15.3、13.5个百分点,性能优于YOLOv7、YOLOv8等模型。在台架试验中,改进后模型检测的含杂率平均相对误差为19.58%,比改进前模型降低了38.12个百分点;含杂率平均值为7.31%,比人工测量的实际含杂率高出0.05个百分点。因此,此方法是一种实时性强、效率高、准确性高且能全量检测机收蔗含杂率的方法,能够为田间甘蔗收获作业质量提供技术支撑。展开更多
基金Project(61101185)supported by the National Natural Science Foundation of China
文摘A fast algorithm based on the grayscale distribution of infrared target and the weighted kernel function was proposed for the moving target detection(MTD) in dynamic scene of image series. This algorithm is used to deal with issues like the large computational complexity, the fluctuation of grayscale, and the noise in infrared images. Four characteristic points were selected by analyzing the grayscale distribution in infrared image, of which the series was quickly matched with an affine transformation model. The image was then divided into 32×32 squares and the gray-weighted kernel(GWK) for each square was calculated. At last, the MTD was carried out according to the variation of the four GWKs. The results indicate that the MTD can be achieved in real time using the algorithm with the fluctuations of grayscale and noise can be effectively suppressed. The detection probability is greater than 90% with the false alarm rate lower than 5% when the calculation time is less than 40 ms.
文摘We prove some value-distribution results for a class of L-functions with rational moving targets. The class contains Selberg class, as well as the Riemann-zeta function.
基金supported by the Ministry of Health and Department of Educational Assistance,University and Research of the Autonomous Province of Bolzano
文摘Parkinson＇s disease （PD） is a progressive neurodegenerative disease, which is generally considered a multifactorial disorder that arises owing to a combination of genes and environmental factors. While most cases are idiopathic, in about 10% of the patients a genetic cause can be detected, ascribable to mutations in more than a dozen genes. PD is characterized clinically by tremor, rigidity, reduced mo- tor activity （bradykinesia）, and postural instability and pathological- ly by loss of dopaminergic （DA） neurons in the substantia nigra pars compacta, loss of DA innervation in the striatum, and the presence of a-synuclein positive aggregates in the form of Lewy bodies. The symptomatic treatment of PD with levodopa, which aims at replac- ing dopamine, remains the gold standard, and no neuroprotective or disease-modifying therapy is available. During treatment, the disease continues to progress, and long-term use of levodopa has import- ant limitations including motor complications termed dyskinesias. Therefore, a pharmacological therapy able to prevent or halt the neu- rodegenerative process is urgently required.
文摘In this paper, a novel signal processing technique hasbeen developed to refocus moving targets image from their smeared responses in the Synthetic Aperture Radar (SAR) image according to the characteristics of the received signals for moving targets. Quadratic Phase Function is introduced to the parameters estimation for moving target echo and SAR imaging. Our method is available even under a low SNR environment and acquiring an exact SAR image of moving targets. The simulated results demonstrated the validity of the algorithm proposed.
文摘针对朝天椒检测深度学习网络模型体积大、参数多以及在计算资源有限的移动设备中难以部署等问题,文中提出一种基于YOLOv5s(You Only Look Once version 5s)的轻量级朝天椒检测模型。利用GhostNet中Ghost模块和Ghost瓶颈(Ghost BottleNeck)结构重新构建特征提取网络,减少模型的参数量和计算复杂度。在特征融合部分采用GSConv(Ghost-Shuffle Convolution)轻量级卷积和VoV-GSCSP(VoVNet-Ghost Shuffle-Cross Stage Partial)结构分别替换原始卷积和CSP(Cross Stage Partial)模块,在保证精度的同时使模型轻量化效果最佳。采用角度惩罚度量的SIoU(SCYLLA-Intersaction over Union)损失优化边界框损失函数,提升了轻量化后的模型精度和泛化能力。实验结果表明,改进YOLOv5s-GGS(YOLOv5s-GhostNet GSConv SIoU)模型相较于原始网络模型的精确度、召回率和平均精度均值(mean Average Precision,mAP)分别提高了7.0百分点、3.5百分点和3.8百分点,参数量、计算复杂度和权重降低了42%以上。相较于主流目标检测模型,所提模型具有更高的检测精度以及更少的模型体积,实现了模型的轻量化,精度提升较大,推理速度较快,更适合部署于移动设备。
文摘切段式机收甘蔗含杂率的自动测量可以客观评估机收甘蔗到糖厂入榨前的质量。针对现有抽样称重估算杂质方式效率低且主观性强的问题,以及因田间环境较为复杂使得检测目标蔗段存在运动状态变换导致的模糊、光照强度变化和蔗叶遮挡等技术难点,提出了一种基于改进YOLOv5安装在切段式甘蔗机上的机收蔗含杂率检测的方法。首先,针对工业相机拍摄的蔗段目标为小目标的应用场景,增加小目标检测层,增强网络模型对其的专注;其次,将C3模块替换成C2f模块,提高网络模型对小物体、低对比度目标的检测速度和检测精度;最后,加入加权交并比WIoU(Weighted Intersection over Union)损失函数,提升预测框的回归精度,增强数据集训练效果。试验结果表明:基于改进YOLOv5的机收蔗含杂率检测模型,蔗段识别准确率达95.2%、mAP(mean Average Precision)值为62.5%,相较于原始YOLOv5模型分别提高了15.3、13.5个百分点,性能优于YOLOv7、YOLOv8等模型。在台架试验中,改进后模型检测的含杂率平均相对误差为19.58%,比改进前模型降低了38.12个百分点;含杂率平均值为7.31%,比人工测量的实际含杂率高出0.05个百分点。因此,此方法是一种实时性强、效率高、准确性高且能全量检测机收蔗含杂率的方法,能够为田间甘蔗收获作业质量提供技术支撑。
