基于无人机视频影像的油菜苗检测与计数

针对油菜生长早期传统人工苗情调查方法效率低、主观意识强,不能满足大面积或经常性高精度苗期调查作业需求的问题,该研究基于无人机影像及机器学习技术,提出一种油菜苗视频流检测模型及计数方法。通过对YOLO系列基础模型添加多头自注意力,用BasicRFB（basic receptive field block）模块替换原有的空间池化结构（spatial pyramid pooling-fast,SPPF）模块,并对Neck部分添加一维卷积及更换下采样方式等,进一步结合DeepSORT（deep simple online and real-time tracking）算法和越线计数技术实现对油菜苗的数量统计。算例测试结果表明,改进后YOLOv5s的交并比阈值0.50的平均精度均值达到93.1%,交并比阈值0.50～0.95的平均精度均值达到了67.5%,明显优于Faster R-CNN、SSD和YOLOX等其他经典目标检测算法,交并比阈值0.50的平均精度均值分别高出14.82、26.37和3.3个百分点,交并比阈值为0.50～0.95范围下的平均精度均值分别高出25.7、33.9和6.7个百分点。油菜苗计数试验结果表明,离线视频计数时,在合理的种植密度区间内,所提算法的油菜苗计数精度平均达到93.75%,计数效率为人工计数的9.54倍;在线实时计数时,在不同天气情况下,计数平台的油菜苗计数精度最大相差1.87个百分点,具有良好的泛化性,满足油菜苗计数实时性要求。     

水下底栖生物轻量化目标检测算法YOLOv7-RFPCW

针对水下底栖生物检测模型参数量过多,计算量过大,同时水下环境复杂容易造成错检和漏检,影响模型在水下底栖生物检测任务中的准确性的问题。提出了一种水下底栖生物轻量化检测算法YOLOv7-RFPCW。对YOLOv7网络重新设计轻量级网络结构,降低了特征提取网络的参数量和计算量,减少模型体积。设计了P-ELAN和P-ELAN-W模块,进一步轻量化特征提取网络;针对水下图像颜色失真,目标的空间位置不准确的问题,采用CBAM注意力模块加强特征融合,减少信息丢失,以更好地适应特殊的水下环境;针对水下目标容易出现形状变形的问题,使用WIOU-V3损失函数替换默认的CIOU损失函数,提高水下底栖生物检测的鲁棒性。试验结果显示,改进后的模型YOLOv7-RFPCW的参数量和计算量分别减少了75.9%和30.7%,模型体积减小了75.3%,精度提升了1.9个百分点。这一综合性的提升兼顾了轻量化和精度,为在水下环境中部署提供了可行的解决方案,验证了所提出的改进算法能胜任水下底栖生物检测任务。     

浅谈大学生合作学习的意义及培养策略

我国大学生未充分利用合作学习的方法,缺少合作精神。可通过创建和谐校园环境和良好的家庭环境,改革以应试为主的教育体制,培养大学生合作学习的意识和社交技能大学生合作学习能力。     

传统学习行为与网络学习行为比较研究

网络媒体的发展使得网络学习行为的研究成为热点。把传统学习行为与网络学习行为进行对比,比较两者的优缺点,找出网络学习的规律和特点,更好地服务于网络教学,提升教学水平。     

关于生物类基础化学教学中应用“问题学习法”的探讨

本文通过课前、课堂、课后三个部分的问题设置,紧扣教学大纲,紧密联系中学化学和后续课程中相关知识,体现理论联系实际、学科前沿。引领学生"主动学习,积极思考",为后续课程学习起到积极的、有效的铺垫作用。     

基于可信多数投票的快速概念漂移检测

数据流因具有数据持续到达，概念漂移产生时刻无法预测、概念的数量不确定等特征，使得滑动窗口的大小很难事先确定，滑动窗口包含概念的数量对概念漂移检测存在影响.本文提出了基于可信多数投票的快速概念漂移检测算法（CMV_SEA），该算法使用SEA算法中的基分类器淘汰方法，使用可信多数投票实现滑动窗口中基分类器的集成.仿真实验表明：相比于SEA算法，CMV_SEA算法提高了泛化能力；能在新概念产生的第一时间内检测到概念漂移；对概念漂移的检测能力和新概念的学习能力不受滑动窗口大小的影响.     

CAI课件与CAI-I课件特点之比较分析

随着计算机技术和网络技术的发展,传统的计算机辅助教学（ＣＡＩ）手段与Ｉｎｔｅｒｎｅｔ技术的有机 结合,产生了ＣＡＩ的新形式—基于Ｉｎｔｅｒｎｅｔ的分布式ＣＡＩ,我国对此研究尚处于初级阶段。主要对 这种新形势下的ＣＡＩ－Ｉ课件与传统的ＣＡＩ课件的特点加以分析和比较,并简要介绍其实现方法。     

Learning performances of honeybees (Apis mellifera L) are differentially affected by imidacloprid according to the season

To establish the sublethal concentrations domain, acute and chronic oral tests were conducted on caged honeybee workers (Apis mellifera L) using imidacloprid and a metabolite, 5-OH-imidacloprid, under laboratory conditions. The latter showed a 48-h oral LD50 value (153 ng per bee) five times higher than that of imidacloprid (30 ng per bee). Chronic feeding tests indicated that the lowest observed effect concentrations (LOEC) of imidacloprid and of 5-OH-imidacloprid on mortality of winter bees were 24 and 120 microg kg(-1) respectively. Behavioural effects of imidacloprid and 5-OH-imidacloprid were studied using the olfactory conditioning of proboscis extension response at two periods of the year. Winter bees surviving chronic treatment with imidacloprid and 5-OH-imidacloprid had reduced learning performances. The LOEC of imidacloprid was lower in summer bees (12 microg kg(-1)) than in winter bees (48 microg kg(-1)), which points to a greater sensitivity of honeybees behaviour in summer bees, compared to winter bees.     

Stimulating effects of the insecticide chlorpyrifos on host searching and infestation efficacy of a parasitoid wasp

Hymenopterous parasitoids play an important role in the control of insect populations. During oviposition, Hymenopterous parasitoids use cues such as odours from their environment to locate their specific host. Leptopilina heterotoma (parasitoid of Drosophila larvae) locate their host by probing the substrate with the ovipositor. This behaviour can be induced by the odour of the host substrate alone. We analysed the sub-lethal effects of chlorpyrifos at LD20 on the probing activity in response to a fruit odour (banana). The insecticide increased the percentage of females spontaneously probing in response to the odour. Parasitoid females were then conditioned to associate banana odour with the oviposition in host larvae. This conditioning enables parasitoids to memorize the odour and to increase their probing response to this odour. During the olfactory conditioning, females exposed to the insecticide found and oviposited in host larvae more quickly than control females. One hour after the olfactory conditioning, females exposed to the insecticide presented a higher increase of their probing response to the odour than controls. Twenty-four hours after conditioning, the stimulation produced by chlorpyrifos was no longer perceptible, but the level of response of conditioned females was still higher than that of non-conditioned females, showing that odour memory was not impaired by the insecticide treatment. These sub-lethal effects, that stimulate host searching by parasitoids without impairing odour memorization, could increase their parasitic efficiency.     

利用显著图构建注意力深度网络检测诱虫板蔬菜害虫

为提高诱虫板图像蔬菜害虫检测精度,针对背景区域容易导致误检的问题基于显著图分析技术构建了一种注意力深度网络害虫智能视觉检测方法。首先通过显著图筛选出粗候选区域;然后在粗候选区域内用全卷积神经网络精选出细候选区域;接着用神经网络分类器识别细候选区域害虫种类,得到含有冗余的若干检测框;最后用改进的非极大值抑制消除冗余检测框,实现诱虫板图像中目标害虫的检测。针对小菜蛾和瓜实蝇展开试验,获得86.40%的平均精度均值和0.111只的平均绝对计数误差均值,所提方法平均精度均值比Faster R-CNN和YOLOv4分别高2.74和1.56个百分点,平均绝对计数误差均值比Faster R-CNN和YOLOv4分别低0.006和0.003只;同时,消融试验中移除显著图注意力模块后平均精度均值下降了4个百分点、平均绝对计数误差均值增加了0.207只。试验结果表明,所提方法有效提高了诱虫板图像蔬菜害虫检测精度,其中,引入显著图注意力模块对提升检测精度有重要作用。