基于区域面积和特征的图像配准

图像配准是图像处理领域的一个研究热点。双目线索是人类和动物的视觉系统获取深度信息的主要手段之一,而配准是视差测距过程中最主要的步骤。本文主要实现了基于面积的配准和基于边缘点特征的Barnard方法,并且对有关的实验结果进行了直观的分析。     

一种新的基于条件数的图像配准算法

提出了一种新的基于条件数的图像配准算法。该方法在Harris算法提取角点的基础上,采用条件数定量地分析了噪声对确定图像间变换关系的影响程度,通过阈值设定筛选出具有良好稳定性的角点,克服了Harris角点检测可能存在的角点位置偏移和易受噪而提取出伪角点等问题。最后选择了Random Sample Consensus(RANSAC)匹配准则来确定匹配点对。经过实验证明了该配准算法具有精确性、抗噪性和鲁棒性。     

基于Dividing Rectangles的多模态

为了准确、可靠地配准多模态医学图像,提出了一种基于互信息的全局优化配准算法。该算法首先提取出目标物体的外轮廓面,再用迭代最近点方法初步对齐图像;然后用确定性的全局优化方法—Dividing Rectangles搜索归一化互信息的全局最优解。该算法利用图像的特征信息,为Dividing Rectangles方法提供了一个较好的初始配准位置,并充分利用了Dividing Rectangles方法在小范围内的高效搜索能力。实验结果表明,对于3维人体脑部数据,该算法配准精度高、速度快,而且有效地避免了配准过程中出现的局部极值。     

基于sMLD特征的远红外航空图像配准算法

航空场景下的远红外图像迫切需要准确、鲁棒、快速的特征描述与自动配准方法。由于已有的多重线型描述符MLD特征存在“特征孤岛”和“尺度变换受限”的问题,因此针对远红外航空图像的配准问题,文章提出了一种聚合特征点以及线型描述符分段统计的特征描述方式,即sMLD特征。结合sMLD特征相互连接构成网状拓扑结构的特性,还提出了一种由粗到细的分支加速匹配算法RF-BA。其中,RF-BA粗匹配充分利用拓扑图的结构,通过局部寻优算法提高匹配的效率。RF-BA精匹配利用最小外接凸四边形原则和GMS校验原则,提升配准精度。实验结果表明,与已有的几种代表性配准方法相比,所提方法在配准精度和时间开销方面均具有更好的性能。     

基于改进ICP的复杂机械零件测量点云配准方法

点云配准是基于机器视觉进行复杂机械零件三维非接触精密测量的关键环节。针对传统迭代最近点(iterative closest point, ICP)算法对初始位置依赖性强,迭代收敛速度慢,错误对应点对多,难以满足大批量复杂机械零件测量点云配准效率和精度要求的问题,提出了一种基于ISS-FPFH(intrinsic shape signature-fast point feature histogram)特征结合改进ICP的复杂机械零件测量点云配准方法。为了减少点云配准数量,并保留点云表面原来的细微特征,提出了基于重心邻近点的体素滤波器对点云进行下采样预处理。为解决传统ICP算法因合适初始位置难以确定而导致多视角测量点云配准失败的问题,采用了基于ISS-FPFH特征的采样一致性初始配准(sample consensus intial alignment, SAC-IA)算法进行粗配准。为解决传统ICP算法迭代收敛速度慢、错误对应点对多的问题,提出结合法向量夹角约束的点到平面ICP算法进行精配准。以斯坦福大学的bunny点云模型为对象,验证了本文提出方法对噪声点云的鲁棒性。以常见的复杂机械零...     

Semantic Plane-Structure based motion detection with a nonstationary camera

This paper presents a novel method to accurately detect moving objects from a video sequence captured using a nonstationary camera. Although common methods provide effective motion detection for static backgrounds or through only planar-perspective transformation, many detection errors occur when the background contains complex dynamic interferences or the camera undergoes unknown motions. To solve this problem, this study proposed a motion detection method that incorporates temporal motion and spatial structure. In the proposed method, first, spatial semantic planes are segmented, and image registration based on stable background planes is applied to overcome the interferences of the foreground and dynamic background. Thus, the estimated dense temporal motion ensures that small moving objects are not missed. Second, motion pixels are mapped on semantic planes, and then, the spatial distribution constraints of motion pixels, regional shapes and plane semantics, which are integrated into a planar structure, are used to minimise false positives. Finally, based on the dense temporal motion and spatial structure, moving objects are accurately detected. The experimental results on CDnet dataset, Pbi dataset, Aeroscapes dataset, and other challenging self-captured videos under difficult conditions, such as fast camera movement, large zoom variation, video jitters, and dynamic background, revealed that the proposed method can remove background movements, dynamic interferences, and marginal noises and can effectively obtain complete moving objects.© 2017 ElsevierInc.Allrightsreserved.     

Soft Robotic-Adapted Multimodal Sensors Derived from Entirely Intrinsic Self-Healing and Stretchable Cross-Linked Networks

Flexible sensing technologies that play a pivotal role in endowing robots with detection capabilities and monitoring their motions are impulsively desired for intelligent robotics systems. However, integrating and constructing reliable and sustainable flexible sensors with multifunctionality for robots remains an everlasting challenge. Herein, an entirely intrinsic self-healing, stretchable, and attachable multimodal sensor is developed that can be conformally integrated with soft robots to identify diverse signals. The dynamic bonds cross-linked networks including the insulating polymer and conductive hydrogel with good comprehensive performances are designed to fabricate the sensor with prolonged lifespan and improved reliability. Benefiting from the self-adhesiveness of the hydrogel, strong interfacial bonding can be formed on various surfaces, which promotes the conformable integration of the sensor with robots. Due to the ionic transportation mechanism, the sensor can detect strain and temperature based on piezoresistive and thermoresistive effect, respectively. Moreover, the sensor can work in triboelectric mode to achieve self-powered sensing. Various information can be identified from the electrical signals generated by the sensor, including hand gestures, soft robot crawling motions, a message of code, the temperature of objects, and the type of materials, holding great promise in the fields of environmental detection, wearable devices, human-machine interfacing, and robotics.     

Multi-Stage Phase-Segregation of Mixed Halide Perovskites under Illumination: A Quantitative Comparison of Experimental Observations and Thermodynamic Models

Photo- and charge-carrier-induced ion migration is a major challenge when utilizing metal halide perovskite semiconductors for optoelectronic applications. For mixed iodide/bromide perovskites, the compositional instability due to light- or electrical bias induced phase-segregation restricts the exploitation of the entire bandgap range. Previous experimental and theoretical work suggests that excited states or charge carriers trigger the process, but the exact mechanism is still under debate. To identify the mechanism and cause of light-induced phase-segregation phenomena, the full compositional range of methylammonium lead bromide/iodide samples are investigated, MAPb(Br_xI_1-x)₃ with x = 0…1, by simultaneous in situ X-ray diffraction (XRD) and photoluminescence (PL) spectroscopy during illumination. The quantitative comparison of composition-dependent in situ XRD and PL shows that at excitation densities of 1 sun, only the initial stage of photo-segregation is rationalized with the previously established thermodynamic models. However, a progression of the phase segregation is observed that is rationalized by considering long-lived accumulative photo-induced material alterations. It is suggested that (additional) photo-induced defects, possibly halide vacancies and interstitials, need to be considered to fully rationalize light-induced phase segregation and anticipate the findings to provide crucial insight for the development of more sophisticated models.     

Petromyzontidae-Biomimetic Multimodal Microneedles-Integrated Bioelectronic Catheters for Theranostic Endoscopic Surgery

Current catheter devices in minimally invasive surgery still possess limited functional options, lacking multimodal integration of both sensing and therapy. Catheter devices usually operate outside the tissue, incapable to detect intra-tissue biochemical information for accurate localization and assessment of lesions during surgery. Inspired by the feature and functions of Petromyzontidae, here a multimodal core-shell microneedles-integrated bioelectronic catheter (MNIBC) for tissue-penetrating theranostics in endoscopic surgery is developed. The microneedle (MN) device possesses individually addressable functionality at single-MN tip resolution, enabling multiplex functions (a total of 11 functions distributed in three types of catheters) including biochemical sensing, myoelectric modulation, electroporation, and drug delivery in a submucosal environment. The MNIBC is prepared through hybrid fabrication and dimensionality reduction strategies, where the MN electrodes are functionalized with an MXene-carbon nanotube (MXene-CNT)-based electron mediator, addressing the challenge of reduced electrode sensitivity on ultra-small MN tip. The functionalities of MNIBC are demonstrated both ex vivo and in vivo on anesthetized rabbits via laparoscopy, simulated cystoscopy, and laparotomy. The MNIBC can effectively detect intra-tissue biochemical signals in the bladder, and offers localized electroporation and intra-tissue drug delivery for precise treatments of lesions. The versatile features of the MNIBC present a highly advanced platform for precise surgeries.     

基于姿态编码器的2D/3D脊椎医学图像实时配准方法

2D/3D医学图像配准是骨科手术三维实时导航中的一项关键技术,然而传统的基于优化迭代的2D/3D配准方法需要经过多次迭代计算,无法满足医生在手术过程中对于实时配准的要求。针对该问题,提出一种基于自编码器的姿态回归网络来通过隐空间解码捕获几何姿态信息,从而快速地回归出术中X射线图像对应的术前脊椎位置的3D姿态,并经过重新投影生成最终的配准图像。通过引入新的损失函数,以“粗细”结合配准的方式对模型进行约束,保证了姿态回归的精确度。在CTSpine1K脊椎数据集中抽取100组CT扫描图像进行10折交叉验证,实验结果表明：所提出的模型所生成的配准结果图像与X射线图像的平均绝对误差（MAE）为0.04,平均目标配准误差（mTRE）为1.16 mm,单帧耗时1.7 s。与基于传统优化的方法相比,该模型配准时间大幅缩短。相较于基于学习的方法,该模型在快速配准的同时,保证了较高的配准精度。可见,所提模型可以满足术中实时高精配准的要求。