ANALYSIS OF TIPOVER STABILITY FOR NOVEL SHAPE SHIFTING MODULAR ROBOT

A novel three-module robot has been introduced. It can change its configuration to adapt to the uneven terrain and to improve its tipover stability. This three-module tracked robot has three kinds of symmetry configuration. They are line type, triangle type, and row type. After the factors and the countermeasures of mobile robot's tipover problem are analyzed, stability pyramid and tipover stability index are proposed to globally determinate the mobile robot's static stability and dynamic stability. The shape shifting robot is tested by this technique under the combined disturbance of pitch, roll and yaw in simulation. The simulation result shows that this technique is effective for the analysis of mobile robot's tipover stability, especially for the reconfigurable or shape shifting modular robot. Experiments on three symmetry configurations are made under unstructured environments. The environment experiment shows the same result as that of the simulation that the triangle type configuration has the best stability. Both simulation and experiment provide a valid reference for the reconfigurable robot's potential application.     

基于多级特征和混合注意力机制的室内人群检测网络

针对室内人群目标尺度和姿态多样性、人头目标易与周围物体特征混淆的问题,提出了一种基于多级特征和混合注意力机制的室内人群检测网络（MFANet）。该网络结构包括三部分,即特征融合模块、多尺度空洞卷积金字塔特征分解模块以及混合注意力模块。首先,通过将浅层特征和中间层特征信息融合,形成包含上下文信息的融合特征,用于解决浅层特征图中小目标语义信息不丰富、分类能力弱的问题;然后,利用空洞卷积增大感受野而不增加参数的特性,对融合特征进行多尺度分解,形成新的小目标检测分支,实现网络对多尺度目标的定位和检测;最后,用局部混合注意力模块来融合全局像素关联空间注意力和通道注意力,增强对关键信息贡献大的特征,来增强网络对目标和背景的区分能力。实验结果表明,所提方法在室内监控场景数据集SCUT-HEAD上达到了0.94的准确率、0.91的召回率和0.92的F1分数,在召回率、准确率和F1指标上均明显优于当前用于室内人群检测的其他算法。     

基于多级特征整合的图像语义分割研究

传统的全卷积神经网络由于不断的池化和下采样操作使得最后的特征热点图分辨率损失严重,导致了分割结果的细节刻画能力的缺失,为了弥补这一缺陷,往往通过跳跃连接融合中层的特征图以恢复空间信息.由于无法充分利用网络的低层特征信息,传统全卷积网络的特征融合阶段存在相当的缺陷,本文对这一现象进行了深入的分析.本文在上采样路径之前采用基于特征金字塔的特征信息增强方法,克服了浅层特征图语义信息匮乏这一缺点,使得整个网络能更充分的利用前向计算产生的特征图,输出的分割结果也更为精确.本文提出的算法在Pascal VOC数据集上取得了75.8%的均像素精度和83.9%的权频交并比,有效的提高了分类精度.     

特征注意金字塔调制网络的视频目标分割

目的 视频目标分割是在给定第1帧标注对象掩模条件下,实现对整个视频序列中感兴趣目标的分割。但是由于分割对象尺度的多样性,现有的视频目标分割算法缺乏有效的策略来融合不同尺度的特征信息。因此,本文提出一种特征注意金字塔调制网络模块用于视频目标分割。方法 首先利用视觉调制器网络和空间调制器网络学习分割对象的视觉和空间信息,并以此为先验引导分割模型适应特定对象的外观。然后通过特征注意金字塔模块挖掘全局上下文信息,解决分割对象多尺度的问题。结果 实验表明,在DAVIS 2016数据集上,本文方法在不使用在线微调的情况下,与使用在线微调的最先进方法相比,表现出更具竞争力的结果,J-mean指标达到了78.7%。在使用在线微调后,本文方法的性能在DAVIS 2017数据集上实现了最好的结果,J-mean指标达到了68.8%。结论 特征注意金字塔调制网络的视频目标分割算法在对感兴趣对象分割的同时,针对不同尺度的对象掩模能有效结合上下文信息,减少细节信息的丢失,实现高质量视频对象分割。     

基于跳跃连接金字塔模型的小目标检测

随着深度学习的发展,目标检测已经获得了较高的精度和效率。但是小目标的检测仍然是一个挑战。小目标检测准确率较低的重要原因是没有充分利用高层特征的语义信息和低层特征的细节信息之间的关系。针对上述问题,本文提出一种基于跳跃连接金字塔模型的小目标检测方法。与其他的目标检测方法不同,本文提出利用跳跃连接金字塔结构来融合多层高层语义特征信息和低层特征图的细节信息。而且为了更好地提取不同尺度物体对应的特征信息,在网络模型中采用不同大小的卷积核和不同步长的空洞卷积来提取全局特征信息。在PASCAL VOC和MS COCO数据集上进行了实验,验证了算法的有效性。     

纳米金属银、铜辅助化学刻蚀制绒金刚线切割多晶硅的研究

以金刚线切割多晶硅为原料,研究不同纳米金属催化剂(银、铜)辅助化学刻蚀对纳米结构引入及多晶硅片表面制绒效果的影响。研究结果表明:不同纳米金属物种诱导刻蚀对硅片表面形貌结构的影响巨大,相比于纳米银辅助刻蚀形成的硅纳米线阵列结构而言,纳米铜辅助刻蚀形成的倒金字塔结构在各方面的性能均比较突出,大面积微尺度的倒金字塔阵列结构可以更完美地融合表面低反射率和钝化不佳之间的矛盾,且硅片表面切割纹去除效果明显。当金属铜辅助化学刻蚀制绒15 min时,倒金字塔结构最规则、均匀,且在300~1 100 nm波段范围内,反射率由原片的41.8%降低至5.8%。同时倒金字塔形貌具有优越的减反效果和去除切割纹能力,使得制绒金刚线切割多晶硅片有望用来制备高效率的太阳能电池。     

Enhanced efficiency of graphene-silicon Schottky junction solar cell through inverted pyramid arrays texturation

Nanostructures of silicon are gradually becoming hot candidate due to outstanding capability for trapping light and improving conversion efficiency of solar cell. In this paper, silicon nanowires (SiNWs) and silicon inverted pyramid arrays (SiIPs) were introduced on surface of Gr-Si solar cell through silver and copper-catalyzed chemical etching, respectively. The effects of SiNWs and SiIPs on carrier lifetime, optical properties and efficiency of Gr-SiNWs and Gr-SiIPs solar cells were systematically analyzed. The results show that the inverted pyramid arrays have more excellent ability for balancing antireflectance loss and surface area enlargement. The power conversion efficiency (PCE) and carrier lifetime of Gr-SiIPs devices respectively increase by 62% and 34% by comparing with that of Gr-SiNWs solar cells. Finally, the Gr-SiIPs cell with PCE of 5.63% was successfully achieved through nitric acid doping. This work proposes a new strategy to introduce the inverted pyramid arrays for improving the performance of Gr-Si solar cells.     

基于图像匹配的透明车灯部件检测

针对车灯总成的完整性检测,提出了基于图像金字塔和归一化互相关函数结合的分层匹配算法。利用直方图均衡化和锐化滤波增强图像对比度以及边缘细节信息,再运用归一化互相关相似度函数实现车灯透明部件的匹配,同时采用金字塔分层来提高图像匹配的速度。通过实验确定了合适的匹配窗口大小、金字塔层数、相似度阈值等,实现了车灯透明部件的安装检测。实验结果表明:该方法对车灯透明部件的成功检测率可达到约95%,检测时间约50 ms,具有较高的准确率和实时性。     

Local-Tetra-Patterns for Face Recognition Encoded on Spatial Pyramid Matching

Face recognition is a big challenge in the research field with a lot of problems like misalignment, illumination changes, pose variations, occlusion, and expressions. Providing a single solution to solve all these problems at a time is a challenging task. We have put some effort to provide a solution to solving all these issues by introducing a face recognition model based on local tetra patterns and spatial pyramid matching. The technique is based on a procedure where the input image is passed through an algorithm that extracts local features by using spatial pyramid matching and max-pooling. Finally, the input image is recognized using a robust kernel representation method using extracted features. The qualitative and quantitative analysis of the proposed method is carried on benchmark image datasets. Experimental results showed that the proposed method performs better in terms of standard performance evaluation parameters as compared to state-of-the-art methods on AR, ORL, LFW, and FERET face recognition datasets.     

基于FCOS神经网络的制动主缸内槽缺陷检测方法

针对主缸内槽缺陷检测存在干扰因素复杂、检测精度低等难点,提出了一种基于全卷积单阶段神经网络(FCOS)的主缸内槽缺陷检测算法。利用特征融合金字塔网络进行特征提取并逐像素预测,得到缺陷种类,实现凹槽缺陷的自动检测。实验结果表明,FCOS网络对制动主缸内槽砂眼、划痕、振刀纹缺陷检测的平均精度均值分别为85.2％、87.5％、90.1％,精确度分别为0.98、0.89、0.95。实验结果与Mask R-CNN网络和Faster R-CNN网络的实验结果进行对比,FCOS网络具有更高的准确率,学习时长大幅度缩短,且满足实时检测要求。