Additively manufactured Ti–Ta–Cu alloys for the next-generation load-bearing implants

Bacterial colonization of orthopedic implants is one of the leading causes of failure and clinical complexities for load-bearing metallic implants. Topical or systemic administration of antibiotics may not offer the most efficient defense against colonization, especially in the case of secondary infection, leading to surgical removal of implants and in some cases even limbs. In this study, laser powder bed fusion was implemented to fabricate Ti3Al2V alloy by a 1:1 weight mixture of CpTi and Ti6Al4V powders. Ti-Tantalum(Ta)–Copper(Cu) alloys were further analyzed by the addition of Ta and Cu into the Ti3Al2V custom alloy. The biological,mechanical, and tribo-biocorrosion properties of Ti3Al2V alloy were evaluated. A 10 wt.% Ta(10Ta) and 3 wt.% Cu(3Cu) were added to the Ti3Al2V alloy to enhance biocompatibility and impart inherent bacterial resistance. Additively manufactured implants were investigated for resistance against Pseudomonas aeruginosa and Staphylococcus aureus strains of bacteria for up to 48 h. A 3 wt.% Cu addition to Ti3Al2V displayed improved antibacterial efficacy, i.e.78%–86% with respect to CpTi. Mechanical properties for Ti3Al2V–10Ta–3Cu alloy were evaluated, demonstrating excellent fatigue resistance, exceptional shear strength, and improved tribological and tribo-biocorrosion characteristics when compared to Ti6Al4V. In vivo studies using a rat distal femur model revealed improved early-stage osseointegration for alloys with10 wt.% Ta addition compared to CpTi and Ti6Al4V. The 3 wt.% Cu-added compositions displayed biocompatibility and no adverse infammatory response in vivo. Our results establish the Ti3Al2V–10Ta–3Cu alloy’s synergistic effect on improving both in vivo biocompatibility and microbial resistance for the next generation of load-bearing metallic implants.

Leveraging Augmented Reality,Semantic-Segmentation,and VANETs for Enhanced Driver’s Safety Assistance

Overtaking is a crucial maneuver in road transportation that requires a clear view of the road ahead.However,limited visibility of ahead vehicles can often make it challenging for drivers to assess the safety of overtaking maneuvers,leading to accidents and fatalities.In this paper,we consider atrous convolution,a powerful tool for explicitly adjusting the field-of-view of a filter as well as controlling the resolution of feature responses generated by Deep Convolutional Neural Networks in the context of semantic image segmentation.This article explores the potential of seeing-through vehicles as a solution to enhance overtaking safety.See-through vehicles leverage advanced technologies such as cameras,sensors,and displays to provide drivers with a real-time view of the vehicle ahead,including the areas hidden from their direct line of sight.To address the problems of safe passing and occlusion by huge vehicles,we designed a see-through vehicle system in this study,we employed a windshield display in the back car together with cameras in both cars.The server within the back car was used to segment the car,and the segmented portion of the car displayed the video from the front car.Our see-through system improves the driver’s field of vision and helps him change lanes,cross a large car that is blocking their view,and safely overtake other vehicles.Our network was trained and tested on the Cityscape dataset using semantic segmentation.This transparent technique will instruct the driver on the concealed traffic situation that the front vehicle has obscured.For our findings,we have achieved 97.1% F1-score.The article also discusses the challenges and opportunities of implementing see-through vehicles in real-world scenarios,including technical,regulatory,and user acceptance factors.

Chemical incorporation of SiO_(2) into TiO_(2) layer by green plasma enhancer and quencher agents for synchronized improvements in the protective and bioactive properties

This study explores the dynamic interaction between environmentally sustainable plasma enhancer and quencher agents during the incorporation of SiO_(2) into a TiO_(2) layer,with the primary objective of simultaneously augmenting protective and bioactive attributes.This enhancement is realized through the synergistic utilization of Tetraethyl orthosilicate(TE)and Stevia(ST)within a plasma-assisted oxidation process.To achieve this goal,Ti–6Al–4V alloy underwent oxidation in an electrolyte solution containing acetate-glycerophosphate,with the addition of TE and ST separately and in combination.TE,as a silicon oxide(SiO_(2))precursor,facilitates the creation of a calcium-rich,rough,porous layer by undergoing hydrolysis to generate silanol groups(Si–OH),which subsequently condense into silicon-oxygen-silicon(Si–O–Si)bonds,resulting in SiO_(2) formation.In contrast,ST acts as a plasma quencher,absorbing highly reactive plasma species during the oxidation process,reducing energy levels,and diminishing sparking intensity.The combination of TE and ST results in moderate sparking,balancing Stevia's quenching effect and TE's sparking influence.As a result,this coating exhibits enhanced corrosion resistance and bioactivity compared to using either ST or TE alone.The study highlights the potential of this synergistic approach for advanced TiO_(2)-based coatings.

Refractive outcomes after V4c Toric collamer lens implantation over 1y of follow-up

AIM:To evaluate refractive outcomes and corneal astigmatism changes after Toric implantable collamer lens with a central port(V4c T-ICL)implantation over 1y of follow-up.METHODS:A retrospective study was performed including 50 eyes of 50 patients that underwent V4c T-ICL implantation.Uncorrected(UDVA)and corrected(CDVA)distance visual acuities,refraction,refractive and corneal astigmatism changes and corneal coupling correction were evaluated preoperatively,1 and 12mo postoperatively.Vector analysis was used for astigmatism changes.Coefficient of adjustment(CAdj)was calculated for corneal coupling analysis.RESULTS:The mean UDVA achieved was 0.03 logMAR at 1mo and remained unchanged throughout the whole follow-up(P=0.193).At the last visit,84%of the eyes achieved a CDVA of 0.00 logMAR or better.Regarding spherical equivalent refraction(SEQ),96%of eyes were ranges of±1.00 D and 84%of them within±0.50 D.Also,94%of eyes had a remaining refractive cylinder within±1.00 D and 78%of them within±0.50 D.Both,SEQ and refractive cylinder,remain stable over the postoperative follow-up(P=1.000 and P=0.660,respectively).In terms of surgically induced astigmatism(SIA),no statistically significant differences were found over the follow-up(P=0.102)and under correction was found with a correction index lower than the unit at each visit.A keratometric astigmatism induced of 0.59±0.53(vector mean:0.26×73º)D was reached at the last visit.No significant changes in terms of corneal astigmatism orientation were reported over post-surgery visits(P=0.129 and P=0.097 at 1 and 12mo respectively).No clinical significance was found for CAdj on with-the-rule astigmatism.No postoperative complications resulting from the surgery were found.CONCLUSION:Refractive outcomes suggest that the V4c T-ICL implantation for correction of myopic astigmatism was satisfactory in terms of effectiveness,safety,and stability during 1y of follow-up.Corneal astigmatism induced by the incision around 0.5 D is achieved according to the remaining refractive cylinder found at one-year post-surgery.Corneal coupling analysis results in no unexpected spherical change.

Are all primary omental infarcts truly idiopathic?Five case reports

BACKGROUND Idiopathic omental infarction(IOI)is challenging to diagnose due to its low incidence and vague symptoms.Its differential diagnosis also poses difficulties because it can mimic many intra-abdominal organ pathologies.Although hypercoagulability and thrombosis are among the causes of omental infarction,venous thromboembolism scanning is rarely performed as an etiological investigation.CASE SUMMARY The medical records of the 5 cases,who had the diagnosis of IOI by computed tomography,were examined.The majority of the patients were male(n=4,80%)and the mean age was 31 years(range:21-38).The patients had no previous abdominal surgery or a history of any chronic disease.The main complaint of all patients was persistent abdominal pain.Omental infarction was detected in all patients with contrast-enhanced computed tomography.Conservative treatment was initially preferred in all patients,but it failed in 1 patient(20%).After discharge,all patients were referred to the hematology department for thrombophilia screening.Only 1 patient applied for thrombophilia screening and was homozygous for methylenetetrahydrofolate reductase(A1298C mutation)and heterozygous for a factor V Leiden mutation.CONCLUSION IOI should be considered in the differential diagnosis in patients presenting with progressive and/or persistent right side abdominal pain.Investigating risk factors such as hypercoagulability in patients with IOI is also important in preventing future conditions related to venous thromboembolism.

Milk fat globule epidermal growth factor 8 alleviates liver injury in severe acute pancreatitis by restoring autophagy flux and inhibiting ferroptosis in hepatocytes

BACKGROUND Liver injury is common in severe acute pancreatitis(SAP).Excessive autophagy often leads to an imbalance of homeostasis in hepatocytes,which induces lipid peroxidation and mitochondrial iron deposition and ultimately leads to ferroptosis.Our previous study found that milk fat globule epidermal growth factor 8(MFG-E8)alleviates acinar cell damage during SAP via binding toαvβ3/5 integrins.MFG-E8 also seems to mitigate pancreatic fibrosis via inhibiting chaperone-mediated autophagy.AIM To speculate whether MFG-E8 could also alleviate SAP induced liver injury by restoring the abnormal autophagy flux.METHODS SAP was induced in mice by 2 hly intraperitoneal injections of 4.0 g/kg L-arginine or 7 hly injections of 50μg/kg cerulein plus lipopolysaccharide.mfge8-knockout mice were used to study the effect of MFG-E8 deficiency on SAPinduced liver injury.Cilengitide,a specificαvβ3/5 integrin inhibitor,was used to investigate the possible mechanism of MFG-E8.RESULTS The results showed that MFG-E8 deficiency aggravated SAP-induced liver injury in mice,enhanced autophagy flux in hepatocyte,and worsened the degree of ferroptosis.Exogenous MFG-E8 reduced SAP-induced liver injury in a dose-dependent manner.Mechanistically,MFG-E8 mitigated excessive autophagy and inhibited ferroptosis in liver cells.Cilengitide abolished MFG-E8’s beneficial effects in SAP-induced liver injury.CONCLUSION MFG-E8 acts as an endogenous protective mediator in SAP-induced liver injury.MFG-E8 alleviates the excessive autophagy and inhibits ferroptosis in hepatocytes by binding to integrinαVβ3/5.

Sensitivity investigation of 100-MeV proton irradiation to SiGe HBT single event effect

The single event effect(SEE) sensitivity of silicon–germanium heterojunction bipolar transistor(Si Ge HBT) irradiated by 100-Me V proton is investigated. The simulation results indicate that the most sensitive position of the Si Ge HBT device is the emitter center, where the protons pass through the larger collector-substrate(CS) junction. Furthermore, in this work the experimental studies are also carried out by using 100-Me V proton. In order to consider the influence of temperature on SEE, both simulation and experiment are conducted at a temperature of 93 K. At a cryogenic temperature, the carrier mobility increases, which leads to higher transient current peaks, but the duration of the current decreases significantly.Notably, at the same proton flux, there is only one single event transient(SET) that occurs at 93 K. Thus, the radiation hard ability of the device increases at cryogenic temperatures. The simulation results are found to be qualitatively consistent with the experimental results of 100-Me V protons. To further evaluate the tolerance of the device, the influence of proton on Si Ge HBT after gamma-ray(^(60)Coγ) irradiation is investigated. As a result, as the cumulative dose increases, the introduction of traps results in a significant reduction in both the peak value and duration of the transient currents.

Non-invasively differentiate non-alcoholic steatohepatitis by visualizing hepatic integrinαvβ3 expression with a targeted molecular imaging modality

BACKGROUND Non-invasive methods to diagnose non-alcoholic steatohepatitis(NASH),an inflammatory subtype of non-alcoholic fatty liver disease(NAFLD),are currently unavailable.AIM To develop an integrinαvβ3-targeted molecular imaging modality to differentiate NASH.METHODS Integrinαvβ3 expression was assessed in Human LO2 hepatocytes Scultured with palmitic and oleic acids(FFA).Hepatic integrinαvβ3 expression was analyzed in rabbits fed a high-fat diet(HFD)and in rats fed a high-fat,high-carbohydrate diet(HFCD).After synthesis,cyclic arginine-glycine-aspartic acid peptide(cRGD)was labeled with gadolinium(Gd)and used as a contrast agent in magnetic resonance imaging(MRI)performed on mice fed with HFCD.RESULTS Integrinαvβ3 was markedly expressed on FFA-cultured hepatocytes,unlike the control hepatocytes.Hepatic integrinαvβ3 expression significantly increased in both HFD-fed rabbits and HFCD-fed rats as simple fatty liver(FL)progressed to steatohepatitis.The distribution of integrinαvβ3 in the liver of NASH cases largely overlapped with albumin-positive staining areas.In comparison to mice with simple FL,the relative liver MRI-T1 signal value at 60 minutes post-injection of Gd-labeled cRGD was significantly increased in mice with steatohepatitis(P<0.05),showing a positive correlation with the NAFLD activity score(r=0.945;P<0.01).Hepatic integrinαvβ3 expression was significantly upregulated during NASH development,with hepatocytes being the primary cells expressing integrinαvβ3.CONCLUSION After using Gd-labeled cRGD as a tracer,NASH was successfully distinguished by visualizing hepatic integrinαvβ3 expression with MRI.

4Cr5Mo2V钢与4Cr5MoSiV1钢组织和性能的对比

采用金相检验、冲击试验、扫描电镜及能谱分析等方法对4Cr5Mo2V钢和4Cr5MoSiV1钢的组织、性能进行对比分析,并用软件模拟计算相图。结果表明:4Cr5Mo2V钢较4Cr5MoSiV1钢的V元素含量降低、Mo元素含量升高,改变了钢中MC相的稳定性;4Cr5Mo2V钢中高熔点MC相液析碳化物的溶解温度和含量较4Cr5MoSiV1钢降低,使4Cr5Mo2V钢的冲击韧性增大。

基于PIV/V3V技术的100 MW熔盐泵内部流动可视化测量

以100 MW等级塔式太阳能光热电站熔盐泵为研究对象,采用PIV/V3V技术对内部流场进行可视化测量,基于目前世界上最先进的体三维(3D3C)测量技术,构建体三维测量系统和测试平台,设计了100 MW等级光热电站熔盐泵缩放0.5倍模型泵。基于自主专利技术的三维体流场测试标定方法,首次得到100 MW熔盐泵叶轮流道内部三维流场结构,并提取了叶轮单流道S2流面体内部和涡量值,通过对叶轮流道内速度分布与涡量分布进行分析,发现在偏工况运行时,叶轮流道内发生了更为明显的分离涡现象,三维可视化测量结果与设计要求吻合。

自动痔套扎术结合小“v”切口在治疗IV期混合痔中的效果观察

观察自动痔套扎术结合小“V”切口在治疗IV期混合痔中的效果,为临床治疗提供依据。方法 从在我院进行混合痔疮疾病治疗的病人中挑选六十人当作本次研究的对象,并且将所有的病人按照平均随机分组的原则划分为两个小组,并且为每个小组的病人提供不同的治疗和护理服务,并且对两组病人的的临床治疗效果以及术后并发症情况进行对比分析。结果 观察组内病人的疾病治疗效果相对于对照组内病人的情况较好,并且术后综合评分相对于对照组病人的表现更为优秀。结论 自动痔套扎术结合小“V”切口治疗IV期混合痔疾病具有良好的实践作用,能够从根本上提高疾病治疗的效果,控制手术过程中的出血量,预防各类并发症情况的发生,适合在临床中加以大范围运用。

基于CRV-YOLO的苹果中心花和边花识别方法

苹果树疏花是果园生产管理中的重要环节。准确高效地识别苹果中心花和边花,是研发智能疏花机器人的前提。针对苹果疏花作业中的实际需求,提出了一种基于CRV-YOLO的苹果中心花和边花识别方法。本文基于YOLO v5s模型进行了如下改进:将C-CoTCSP结构融入Backbone,更好地学习上下文信息并提高了模型特征提取能力,提高了模型对外形相似和位置关系不明显的中心花和边花的检测性能。在Backbone中添加改进RFB结构,扩大特征提取感受野并对分支贡献度进行加权,更好地利用了不同尺度特征。采用VariFocal Loss损失函数,提高了模型对遮挡等场景下难识别样本检测能力。在3个品种1837幅图像数据集上进行了实验,结果表明,CRV-YOLO的精确率、召回率和平均精度均值分别为95.6%、92.9%和96.9%,与原模型相比,分别提高3.7、4.3、3.9个百分点,模型受光照变化和苹果品种影响较小。与Faster R-CNN、SSD、YOLOX、YOLO v7模型相比,CRV-YOLO的精确率、平均精度均值、模型内存占用量和复杂度性能最优,召回率接近最优。研究成果可为苹果智能疏花提供技术支持。

基于改进YOLO v5的复杂环境下桑树枝干识别定位方法

为实现复杂自然环境下对桑树嫩叶处枝干的识别检测,改变当前桑叶采摘设备作业过程中依赖人工辅助定位的现状,解决识别目标姿态多样和环境复杂导致的低识别率问题,提出一种基于改进YOLO v5模型的桑树枝干识别模型(YOLO v5-mulberry),并结合深度相机构建定位系统。首先,在YOLO v5的骨干网络中加入CBAM(Convolutional block attention module)注意力机制,提高神经网络对桑树枝干的关注度;并增加小目标层使模型可检测4像素×4像素的目标,提高了模型检测小目标的性能;同时使用GIoU损失函数替换原始网络中的IoU损失函数,有效防止了预测框和真实框尺寸较小时无法正确反映预测框及真实框之间位置关系的情况;随后,完成深度图和彩色图的像素对齐,通过坐标系转换获取桑树枝干三维坐标。试验结果表明:YOLO v5-mulberry检测模型的平均精度均值为94.2%,较原模型提高16.9个百分点,置信度也提高12.1%;模型室外检测时应检测目标数53,实际检测目标数为48,检测率为90.57%;桑树嫩叶处枝干三维坐标识别定位系统的定位误差为(9.4985 mm,11.285 mm,19.11 mm),满足使用要求。该研究可实现桑树嫩叶处枝干的识别与定位,有助于推动桑叶智能化采摘机器人研究。

基于改进COF-YOLO v8n的油茶果静态与动态检测计数方法

针对自然环境下油茶果存在严重遮挡、近景色、小目标等现象,使用YOLO网络存在检测精度低、漏检现象严重等问题,提出对YOLO v8n网络进行改进。首先使用MPDIOU作为YOLO v8n的损失函数,有效解决因为果实重叠导致的漏检问题;其次调整网络,向其中加入小目标检测层,使网络能够关注小目标油茶以及被树叶遮挡的油茶;最后使用SCConv作为特征提取网络,既能兼顾检测精度又能兼顾检测速度。改进COF-YOLO v8n网络精确率、召回率、平均精度均值分别达到97.7%、97%、99%,比未改进的YOLO v8n分别提高3.2、4.8、2.4个百分点,其中严重遮挡情况下油茶检测精确率、召回率、平均精度均值分别达到95.9%、95%、98.5%,分别比YOLO v8n提高4.0、9.1、4.6个百分点。因此改进后COF-YOLO v8n网络能够明显提高油茶在严重遮挡、近景色、小目标均存在情况下的识别精度,减小油茶的漏检。此外,模型能够实现动、静态输入条件下油茶果计数。动态计数借鉴DeepSORT算法的多目标跟踪思想,将改进后COF-YOLO v8n的识别输出作为DeepSORT的输入,实现油茶果实的追踪计数。所得改进模型具有很好的鲁棒性,且模型简单可以嵌入到边缘设备中,不仅可用于指导自动化采收,还可用于果园产量估计,为果园物流分配提供可靠借鉴。

基于YOLO v8n-seg和改进Strongsort的多目标小鼠跟踪方法

多目标小鼠跟踪是小鼠行为分析的基本任务,是研究社交行为的重要方法。针对传统小鼠跟踪方法存在只能跟踪单只小鼠以及对多目标小鼠跟踪需要对小鼠进行标记从而影响小鼠行为等问题,提出了一种基于实例分割网络YOLO v8n-seg和改进Strongsort相结合的多目标小鼠无标记跟踪方法。使用RGB摄像头采集多目标小鼠的日常行为视频,标注小鼠身体部位分割数据集,对数据集进行增强后训练YOLO v8n-seg实例分割网络,经过测试,模型精确率为97.7%,召回率为98.2%,mAP50为99.2%,单幅图像检测时间为3.5 ms,实现了对小鼠身体部位准确且快速地分割,可以满足Strongsort多目标跟踪算法的检测要求。针对Strongsort算法在多目标小鼠跟踪中存在的跟踪错误问题,对Strongsort做了两点改进:对匹配流程进行改进,将未匹配上目标的轨迹和未匹配上轨迹的目标按欧氏距离进行再次匹配;对卡尔曼滤波进行改进,将卡尔曼滤波中表示小鼠位置和运动状态的小鼠身体轮廓外接矩形框替换为以小鼠身体轮廓质心为中心、对角线为小鼠体宽的正方形框。经测试,改进后Strongsort算法的ID跳变数为14,MOTA为97.698%,IDF1为85.435%,MOTP为75.858%,与原Strongsort相比,ID跳变数减少88%,MOTA提升3.266个百分点,IDF1提升27.778个百分点,与Deepsort、ByteTrack和Ocsort相比,在MOTA和IDF1上均有显著提升,且ID跳变数大幅降低,结果表明改进Strongsort算法可以提高多目标无标记小鼠跟踪的稳定性和准确性,为小鼠社交行为分析提供了一种新的技术途径。

基于改进YOLO v7轻量化模型的自然果园环境下苹果识别方法

针对自然果园环境下苹果果实识别中,传统的目标检测算法往往很难在检测模型的检测精度、速度和轻量化方面实现平衡,提出了一种基于改进YOLO v7的轻量化苹果检测模型。首先,引入部分卷积(Partial convolution, PConv)替换多分支堆叠模块中的部分常规卷积进行轻量化改进,以降低模型的参数量和计算量;其次,添加轻量化的高效通道注意力(Efficient channel attention, ECA)模块以提高网络的特征提取能力,改善复杂环境下遮挡目标的错检漏检问题;在模型训练过程中采用基于麻雀搜索算法(Sparrow search algorithm, SSA)的学习率优化策略来进一步提高模型的检测精度。试验结果显示:相比于YOLO v7原始模型,改进后模型的精确率、召回率和平均精度分别提高4.15、0.38、1.39个百分点,其参数量和计算量分别降低22.93%和27.41%,在GPU和CPU上检测单幅图像的平均用时分别减少0.003 s和0.014 s。结果表明,改进后的模型可以实时准确地识别复杂果园环境中的苹果,模型参数量和计算量较小,适合部署于苹果采摘机器人的嵌入式设备上,为实现苹果的无人化智能采摘奠定了基础。

柑橘木虱YOLO v8-MC识别算法与虫情远程监测系统研究

柑橘木虱是黄龙病的主要传播媒介,其发生与活动可对柑橘果园造成毁灭性后果。为实现木虱虫情的高效监测,设计了一种集诱捕拍照、耗材更新、害虫识别与结果展示于一体的智能监测系统。设计了具备诱虫胶带自动更新、虫情图像实时获取功能的诱捕监测装置;应用选点裁剪、Mosaic数据增强(Mosaic data augmentation,MDA)和CA(Coordinate attention)注意力机制,改进了YOLO v8木虱识别模型;开发了Web和手机APP客户端,可实现虫情数据的可视化展示与远程控制。模型测试阶段,改进后的YOLO v8-MC召回率、F1值及精确率分别达到91.20%、91%、90.60%,较基准模型分别提升5.47、5、4.64个百分点;迁移试验中,模型召回率、F1值及精确率分别达到88.64%、87%、84.78%,且系统工作状态良好,满足野外使用需求。开发的智能监测系统能有效实现果园木虱虫情的远程监测,可为此类虫害防治管理提供有效手段。

基于YOLO v5的直播油菜激光间苗系统设计与试验

间苗是保证直播油菜增产的一项关键技术措施,为解决人工间苗劳动强度大、机械间苗不精确的问题,采用机器视觉的方式,基于深度学习算法YOLO v5平台,设计并搭建自动间苗装置。机器视觉系统评估幼苗种群的合理密植情况,间苗算法以间距和幼苗叶展为评估标准,实现控制幼苗间距和筛选优质苗的功能。选用遗传算法对间苗行进路径进行规划,相较于未规划路径可缩短最低为50%的行进距离,最终采用激光器高温烧除的方式完成间苗作业。选取油菜苗作为试验材料,间苗阈值α是划定幼苗最小间距的参数,设置不同的间苗阈值α进行试验。结果表明,间除苗的数量随着间苗阈值α的增加而增加,幼苗平均密度下降的同时种群分布趋于均匀,对间除苗的叶展长度分类统计,α为0~75 mm时,间除苗叶展长度全部在0~20 mm范围;α为75~200 mm时,间除苗叶展长度为0~40 mm,其中叶展长度为20~40 mm的最高占比约为76%;α为200~350 mm时,间除苗叶展长度在40 mm以上的幼苗开始增加,最高占比约为14%,间除苗叶展长度梯次分布证明了间苗算法具备筛选优质苗的性能。间苗执行阶段耗时占据间苗作业总耗时的90%以上,以激光走线参数L、激光器功率P、间苗距离阈值为试验因素,三因素三水平正交试验结果表明:选择合适的激光走线参数L能有效提高间苗死亡率、降低间苗误伤率和减少间苗耗时,在参数L为30 mm、P为7.5 W、α为250 mm下开展土槽台架性能验证试验,激光间苗平均死亡率为93.29%,平均误伤率为5.19%,平均总耗时为15.19 min,为开发基于机器视觉的激光自动间苗机提供了理论基础和技术支撑。

基于改进YOLO v5的复杂环境下花椒簇识别与定位方法

花椒树产果量大,枝干纵横交错,树叶茂密,给花椒的自动化采摘带来了困难。因此,本文设计一种基于改进YOLO v5的复杂环境下花椒簇的快速识别与定位方法。通过在主干提取网络CSPDarknet的CSPLayer层和Neck的上采样之后增加高效通道注意力ECA(Efficient channel attention)来简化CSPLayer层的计算量,提升了特征提取能力。同时在下采样层增加协同注意力机制CA(Coordinate attention),减少下采样过程中信息的损失,强化特征空间信息,配合热力图(Grad-CAM)和点云深度图,来完成花椒簇的空间定位。测试结果表明,与原YOLO v5相比较,改进的网络将残差计算减少至1次,保证了模型轻量化,提升了效率。同帧数区间下,改进后的网络精度为96.27%,对比3个同类特征提取网络YOLO v5、YOLO v5-tiny、Faster R-CNN,改进后网络精确度P分别提升5.37、3.35、15.37个百分点,连株花椒簇的分离识别能力也有较大提升。实验结果表明,自然环境下系统平均识别率为81.60%、漏检率为18.39%,能够满足花椒簇识别要求,为移动端部署创造了条件。

基于YOLO v5的农田杂草识别轻量化方法研究

针对已有杂草识别模型对复杂农田环境下多种目标杂草的识别率低、模型内存占用量大、参数多、识别速度慢等问题,提出了基于YOLO v5的轻量化杂草识别方法。利用带色彩恢复的多尺度视网膜(Multi-scale retinex with color restoration, MSRCR)增强算法对部分图像数据进行预处理,提高边缘细节模糊的图像清晰度,降低图像中的阴影干扰。使用轻量级网络PP-LCNet重置了识别模型中的特征提取网络,减少模型参数量。采用Ghost卷积模块轻量化特征融合网络,进一步降低计算量。为了弥补轻量化造成的模型性能损耗,在特征融合网络末端添加基于标准化的注意力模块(Normalization-based attention module, NAM),增强模型对杂草和玉米幼苗的特征提取能力。此外,通过优化主干网络注意力机制的激活函数来提高模型的非线性拟合能力。在自建数据集上进行实验,实验结果显示,与当前主流目标检测算法YOLO v5s以及成熟的轻量化目标检测算法MobileNet v3-YOLO v5s、ShuffleNet v2-YOLO v5s比较,轻量化后杂草识别模型内存占用量为6.23 MB,分别缩小54.5%、12%和18%;平均精度均值(Mean average precision, mAP)为97.8%,分别提高1.3、5.1、4.4个百分点。单幅图像检测时间为118.1 ms,达到了轻量化要求。在保持较高模型识别精度的同时大幅降低了模型复杂度,可为采用资源有限的移动端设备进行农田杂草识别提供技术支持。