最新刊期
卷 32 ， 期 8 ， 2024
- “旋转激光扫描系统是一种空间多目标并行角度传感系统，其光源性能对系统精度至关重要。为解决光源导入问题，研究人员创新性地设计了一种采用光纤从外部导入光源的新型轴系结构。然而，激光在穿过转轴进入镜组时，倾斜入射柱镜会导致扫描光面型变化。为解决这一难题，研究人员提出了一种基于平行线传递性的激光光轴姿态调整方法，并建立了光轴倾斜误差的标定方法。实验结果证明，经过调整后，光轴与转轴轴线的空间夹角小于0.15°，满足了扫描光截面中心直线度对激光入射柱镜的角度要求。这一研究成果不仅提高了系统的测量精度，也为旋转激光扫描系统在工业测量等领域的应用提供了有力支持。”摘要：The rotating laser-scanning system is a spatial multi-target parallel angle sensing system， which is based on a precision turntable and combined with multiple lasers. To improve the performance of the system's light source， a new type of axis structure was designed that uses optical fibers to import light sources from the outside. However， when the laser passes through the hollow rotating shaft and enters the reflector group at a certain angle， if the incident cylindrical lens is tilted， the scanning light surface pattern will change.To address this problem， a method for adjusting the attitude of the laser optical axis based on the transmissivity of parallel lines was studied. With the help of a laser tracking measurement platform， individual fittings were performed for the optical axis and the rotational axis system. A calibration method for the tilt error of the optical axis was established.The experimental results demonstrate that the spatial angle between the adjusted optical axis and the rotational axis exceeds 0.15°， fully satisfying the angular requirements for laser incidence on the cylindrical lens in terms of the central linearity of the light cross-section.This adjustment scheme is applicable to the assembly phase of the system， enabling the scanned light to approach an ideal plane more closely. This contributes to the enhancement of the system's measurement accuracy.关键词：Large-scale measurement;rotating laser-scanning system;optical axis adjustment;axis alignment;error analysis
- 185
- |
- 35
- |
- 0
发布时间：2024-05-08 - “纳秒激光脉冲技术取得重要突破，研究团队成功实现了高速转镜调Q控制系统，为获得窄脉冲、高峰值功率输出激光提供了直接方式。该系统以Arduino mega 2560单片机为核心，精准控制激光电源的充放电和高速电机启停，同时通过整合降频控制氙灯放电时刻，实现了对延迟时间的精准控制。在5 Hz重复频率下，转镜转速为650 r/s时，获得的最高单脉冲激光能量为45.7 mJ、脉冲宽度为86.2 ns，相应的峰值功率为530.2 kW。这一成果为纳秒激光脉冲技术的进一步发展奠定了基础，有望为激光加工、医疗等领域带来更高效、更精准的解决方案。”摘要：Rotating mirror Q-switched technology incurs no intracavity insertion loss， making it a direct method for obtaining narrow pulses and high peak power output lasers. Nanosecond pulses require the use of high-speed rotating mirror Q-switch， along with precise control of motor speed and xenon lamp discharge delay. This ensures maximum population inversion of in the laser medium， leading to the highest laser energy output. This paper presented the design of a high-speed rotating mirror Q-switched control system with the Arduino Mega 2560 microcontroller as its core. The system utilized precise parse of serial screen instructions by the microcontroller to control the charging and discharging of the laser power supply and the start-stop operations of the high-speed motor. Additionally， it integrated and downshifted the pulse signals from the rotating mirror to control the precise delay time for triggering the xenon lamp discharge， thereby achieving precise control over the delay time for the nanosecond narrow pulse Q-switched output of the lamp-pumped Er，Cr：YSGG laser. The highest single-pulse laser energy of 45.7 mJ， pulse width of 86.2 ns， and corresponding peak power of 530.2 kW were obtained at a repetition frequency of 5 Hz with a rotating mirror speed of 650 r/s.关键词：rotating mirror Q-switch;Solid state laser;Er，Cr：YSGG laser;Arduino;Xe flash lamp
- 82
- |
- 16
- |
- 0
发布时间：2024-05-08 - “齿轮齿面形貌激光干涉测量领域迎来重要突破。以往，由于噪声干扰，采集的包裹相位图中常出现质量差的区域，导致解包路径错误，严重影响测量精度。针对这一问题，研究团队提出了一种基于Tsallis相对熵质量评价的包裹相位修补方法。该方法通过高斯分布拟合阈值处理后的图像，并利用Tsallis相对熵评价相位质量，最后修补质量差的区域，显著提高相位解包裹质量。实验验证显示，该方法可减少40%的相位残差点，有效提升相位质量。这一方法简单易行，准确率高，对其他移相光学领域同样适用，为齿轮齿面形貌激光干涉测量精度提升开辟了新的途径。”摘要：In the laser interferometric measurement of gear tooth flank topography， the collected wrapped phase map can be influenced by various sources of noise. These areas of poor phase quality in the wrapped phase map can lead to unwrapping errors and reduces measurement accuracy. This paper proposed a wrapped-phase repair method to address the problem. It utilized Tsallis relative entropy as an evaluation criterion， based on the probability distribution characteristics of the interference fringe image. The method first fitted the thresholded image using Gaussian distribution， then evaluated the phase quality by Tsallis relative entropy， finally patched the poor quality regions to improve the quality of phase unwrapped. The experimental results show that the method can reduce the phase residue points by 40% and effectively improve the phase quality. The proposed method is user-friendly and highly accurate. It can be applied to other fields related to phase-shifting optics.关键词：interferometry;wrapped phase map;Tsallis relative entropy;image repair;phase unwrapping
- 65
- |
- 21
- |
- 0
发布时间：2024-05-08
Modern Applied Optics
-
Formation mechanism of Poisson's burr in micro-grooved diamond cutting of nickel-phosphorus alloy AI导读
“在金属切削领域，镍磷合金的超精密切削过程中泊松毛刺的形成一直是一个研究热点。为了深入了解其形成机理和规律，并寻找更精准的毛刺尺寸表征方式，科研人员进行了深入研究。他们首先对泊松毛刺的形成机理进行了理论分析，建立了预测模型，并通过实验验证了模型的准确性。实验结果表明，切削深度对毛刺尺寸的影响显著，而切削速度的影响较小。此外，研究还发现，传统的毛刺高度和宽度表征方法存在一定误差，因此科研人员提出了一种基于积分法的泊松毛刺体积计算方法，进一步提高了毛刺尺寸表征的准确性。这一研究成果为镍磷合金正交切削过程中切削参数的合理选择和毛刺尺寸评估提供了重要的理论指导，有望推动金属切削技术的进一步发展。”摘要：In order to explore the formation causes and rules of Poisson's burr in ultra precision cutting of nickel-phosphorus alloy， and to seek a more accurate characterization method of Poisson's burr size， this paper conducted theoretical analysis of the formation mechanism of Poisson's burr in nickel-phosphorus alloy and established a prediction model for the height and width of Poisson's burr in orthogonal cutting process. Through experimental analysis of the influence of cutting parameters on the height and width of Poisson’s burr， it can be observed that the effect of cutting depth on burr size is significant. When the cutting depth increases from 3- 69
- |
- 19
- |
- 0
发布时间：2024-05-08 -
Adaptive precision image stabilization control closed-loop method for space astronomical telescope AI导读
“科技新闻播报：空间天文望远镜领域迎来重大突破。针对低频段视轴扰动补偿问题，专家提出了一种基于主动光学技术的自适应精密稳像闭环控制方法。该方法利用精细导星仪作为高精度视轴扰动检测器，结合四点支撑压电驱动大口径快摆镜机构进行扰动补偿。通过精密稳像闭环控制，该方法能准确计算出补偿视轴扰动所需的摆动角度，并转换为各支撑点压电陶瓷执行器的伸缩量。同时，采用基于广义Bouc-Wen逆模型的压电动态迟滞前馈补偿方法，实现高精度定位控制。此外，该方法还利用具有自学习和自适应能力的单神经元对PID控制器参数进行调优，确保最优控制效果。实验表明，该方法能有效补偿空间天文望远镜的视轴偏差，将精细导星仪X和Y方向星点质心位置偏差功率谱密度在0~6 Hz频段内积分值分别抑制了98.54%和98.62%。这一成果为空间天文望远镜的精确观测提供了有力支持，标志着空间光学控制技术的又一重要进步。”摘要：An adaptive closed-loop control method for precision image stabilization based on active optical technology was proposed to compensate for low-frequency LOS disturbances in space astronomical telescopes. The fine guide sensor （FGS） was used as a high-precision LOS disturbance detector， and the four-point supporting large-aperture fast steering mirror （FSM） mechanism drove by piezoelectric actuators（PZT） was used as a LOS disturbance compensator in this method. First， a PID controller was connected in series with an integral link for precise image stabilization closed-loop control to obtain the two-dimensional swing angles of the FSM required to compensate for the two-dimensional LOS disturbance detected by FGS. Furthermore， the expansion amount of each PZT was calculated through driving structure. Then， the feedforward compensation method based on generalized Bouc-Wen inverse hysteresis model was used for high-precision positioning control of the piezoelectric ceramic actuator. Finally， according to supervised Hebb learning rules， the single neuron with self-learning and adaptive abilities was used to adjust the PID controller parameters， thereby obtaining the optimal controller parameters. The experimental results show that the proposed control method can effectively compensate the LOS disturbance of the space astronomical telescope， and the integral value of PSD of the position error of the star point centroid in the X direction and Y direction of the FGS can be suppressed by 98.54% and 98.62% respectively in the frequency band of 0~6 Hz.关键词：space astronomical telescope;precision image stabilization;fine guide sensor;fast steering mirror;hysteresis compensation;single-neuron adaptive PID control- 70
- |
- 18
- |
- 0
发布时间：2024-05-08 - “在主动磁轴承技术领域，为应对基座运动导致转子振动增大的挑战，研究者提出了一种基于惯性运动的前馈控制方法。通过构建五自由度磁轴承转子动力学模型，并深入分析基座小幅度复杂运动对转子受扰动力的影响，他们设计出了一种自适应惯性前馈控制策略。实验验证表明，这一方法显著降低了磁轴承转子的位移，减少了80%以上，极大提升了磁轴承转子的运行精度。这一创新不仅为磁悬浮系统基座扰动问题提供了有效解决方案，而且其实用性也得到了实验平台的验证，仅需添加一个小型惯性微机电测量单元，即可满足实际工程需求，为磁轴承技术的进一步发展奠定了坚实基础。”摘要：The rotor vibrations can be induced by the base motion in the active magnetic bearing systems. To solve this problem， a feedforward control approach based on inertial motion compensation was devised and implemented. First， a comprehensive five degrees of freedom dynamic model was formulated to describe the dynamics of a magnetically suspended rotor in the active magnetic bearing system with the base movement. Then， the rotor dynamics with various disturbance forces during small-amplitude complex base motions were analyzed. Subsequently， an innovative inertial feedforward method employing an adaptive algorithm was proposed. Finally， to verify the effectiveness of the proposed control method， an experiment platform was built， and then experimental investigations were carried out to compare the rotor's response to various disturbances both before and after activating the feedforward controller. The experimental results show that the implementation of the proposed feedforward control method led to an about 80% reduction of the vibration displacement of the magnetically suspended rotor when it is subjected to base motion perturbations. This marked reduction in displacement significantly enhanced the operational precision of the magnetically suspended rotor. Furthermore， the hardware implementation of the feedforward control method only need the addition of a compact inertial micro-electromechanical measurement unit. This small hardware addition meets some requirements for engineering applications， especially where there is very small space for the mechanical structure. In conclusion， the proposed inertial motion feedforward control method demonstrates promising ability in effectively reducing the vibration displacements of the magnetically suspended rotor disturbed by the base motion， and this can improve the operational stability and precision of active magnetic bearing systems while only a very small inertial measurement unit is added.关键词：active magnetic bearing;moving base compensation;adaptive algorithm;feedforward controller;inertial motion;Vibration suppression
- 56
- |
- 11
- |
- 0
发布时间：2024-05-06
Micro/Nano Technology and Fine Mechanics
- “输电线路弧垂测量领域迎来新的突破。传统的弧垂测量方法操作繁琐且耗时，而单目视觉方法又受限于视场角和分辨率，难以应对大档距输电线路的测量挑战。针对这一问题，研究人员提出了一种结合IMU加速度计传感器和高分辨率工业单目相机的局部摄影测量方法。在已知少量输电线路参数的前提下，只需拍摄待测导线的左侧部分，即可通过引入语义分割技术提取图像中的导线，建立摄影测量模型，进而恢复导线的三维形状并计算弧垂。这一方法不仅操作便捷，而且测量精度较高，相对误差可控制在5%以内，为输电线路弧垂测量提供了新的解决方案，对于保障输电线路的稳定运行具有重要意义。”摘要：Regular monitoring of the size of wire sag in transmission lines is of great significance for the stable operation of transmission lines. The traditional sag measurement method is complex and time-consuming， and obtaining a complete wire image using monocular vision is a convenient sag measurement method. However， when the span of the transmission line is too large， the camera is limited by the field of view angle and resolution， making it difficult to capture the complete wire and measure the sag. To address the above issues， a convenient and usable local photogrammetry method was designed， which combines the accelerometer sensor in the IMU with a high-resolution industrial monocular camera. With a small amount of transmission line parameters known， only the left part of the tested wire needs to be photographed. After obtaining a single wire image， semantic segmentation technology was introduced to extract the wires in the image. Based on the photogrammetric model， the actual points of the wires were connected to the corresponding imaging points， and a nonlinear equation system was established. After nonlinear solving， the three-dimensional shape of the wires was restored， and then the wire sag was calculated. The relative error of sag measurement can be controlled within 5%， meeting practical application requirements.关键词：transmission line;sag;localized photography;Inertial Measurement Unit（IMU）;semantic segmentation
- 50
- |
- 15
- |
- 0
发布时间：2024-05-08 - “针对气体泄漏事故危害大的问题，研究人员在红外成像探测技术领域取得了新突破。传统的气体红外图像对比度低、边缘模糊，识别难度大。为此，专家提出了一种基于混合高斯背景建模的低对比度气体红外图像泄漏检测方法。该方法首先在预处理阶段使用时域自适应帧间滤波算法进行降噪和细节保持，随后结合空域信息和梯度信息约束，构建时空混合高斯背景模型，初步提取泄漏气体目标前景。为进一步提高准确性，研究人员还利用改进的快速鲁棒的模糊C均值聚类方法，实现气体区域的自适应分割。实验结果显示，在5米泄漏距离下，该检测算法准确率高达92.3%~96.3%，有效弥补了气体区域空洞问题，降低了其他运动物体的干扰。这一研究成果不仅显著提升了气体泄漏检测的抗干扰能力和区域分割精度，也为红外成像探测技术的发展和应用提供了新的思路和方向。”摘要：The hazards caused by gas leakage accident are multifaceted， such as environmental pollution， personnel and property loss， fire and explosion. Thermal infrared imaging is widely used as a qualitative detection technology that can realize large-scale and fast imaging. However， compared with general infrared image， the contrast of gas cloud infrared image is lower， the edge is more blurred， and it’s hard to detection. To solve this problem， this article proposed a leak detection method for low contrast gas infrared images based on mixed Gaussian background modeling. Firstly， in the preprocessing stage， time-domain adaptive interframe filtering algorithm was proposed to realize noise reduction and detail maintenance of infrared images. Then， based on spatial information and gradient information constraints， a spatiotemporal mixed Gaussian background model was proposed to achieve preliminary extraction of the foreground of leaked gas targets. Finally， to better remove interfering moving targets in foreground detection， an improved fast and robust fuzzy C-means clustering method was used to realize adaptive segmentation of gas regions. The experimental results show that at the leakage distance of 5 m， this detection algorithm can effectively improve accuracy， compensate for the problems of gas region voids， and reduce interference from other moving objects. The accuracy of gas leakage detection is between 92.3% and 96.3%， which has significant anti-interference and region segmentation capabilities compared to other algorithms.关键词：gas leakage detection;infrared thermal imaging;spatiotemporal Gaussian mixture model;time-domain adaptive inter frame filtering;moving detection;fast and robust fuzzy C-means clustering
- 60
- |
- 19
- |
- 0
发布时间：2024-05-08 - “地质评估领域迎来了一项重要突破。针对砂粒显微图像分割这一难题，研究者提出了一种增强型金枪鱼群优化指数熵的分割方法（ETSO-EXP）。该方法能够有效保留各类砂粒的纹理特征，为地质评估提供了更为准确的图像分割手段。研究者首先针对金枪鱼群优化算法（TSO）的不足，提出了混沌扰动策略、动态权重策略和余弦干扰策略进行增强。实验表明，这些改进使得ETSO在收敛精度和速度上都有了显著提升。随后，研究者将ETSO应用于确定EXP的分割阈值，并通过信息量标准验证了该方案的可行性。在雅鲁藏布江砂粒显微图像数据集上的实验结果显示，与TSO-EXP相比，ETSO-EXP在峰值信噪比、结构相似性、特征相似度和寻优速度等方面均取得了显著的提升。这一研究成果不仅展示了ETSO-EXP在砂粒显微图像分割中的优越性能，还为地质评估领域提供了新的解决方案。该方法对于处理对比度较高、纹理丰富或砂粒碎屑尺寸差异较大的图像具有较高的分割精度和计算速度，为地质评估提供了有力的技术支持。”摘要：Microscopic image segmentation of sand grains can assist geological assessment， but it poses challenges to the accuracy of segmentation due to its variety and complex features. For such images， a segmentation method with enhanced tuna swarm optimization exponential entropy （ETSO-EXP） was proposed， which could effectively preserve the texture features of various sand grains. First of all， aiming at some deficiencies of the tuna swarm optimization （TSO） algorithm in global search and local development， a chaotic disturbance strategy， a dynamic weight strategy and a cosine disturbance strategy were proposed to enhance it. The benchmark function experiment showed that the ETSO greatly improved the convergence accuracy and slightly increased the convergence speed. Secondly， the ETSO algorithm was used to determine the segmentation threshold of the EXP， and the feasibility of the scheme was verified by taking the information content of the segmented image as the standard. Finally， a segmentation experiment was carried out on the Yarlung Zangbo River sand microscopic image dataset. Compared with the TSO-EXP， the image of the ETSO-EXP segmentation has a better peak signal-to-noise ratio， structural similarity， feature similarity and the optimization speed has been improved by 18.78%， 6.85%， 4.16% and 3.83%， respectively， and the performance is the best among the similar segmentation methods. The results show that the segmentation method with the ETSO-EXP has high segmentation accuracy and calculation speed for images with high contrast， rich texture or large differences in the size of sand debris.关键词：sand grain microscopic image;image segmentation;exponential entropy;tuna swarm optimization
- 39
- |
- 23
- |
- 0
发布时间：2024-05-08 - “随着自动驾驶技术的日新月异，城市街景语义分割技术作为自动驾驶技术中的关键环节，正逐渐受到关注。一项新的研究成果在实时城市街景语义分割领域取得了显著突破。研究团队针对当前语义分割算法面临的挑战，如像素区分不清晰、复杂场景理解不精准以及小尺度对象分割不准确等问题，提出了一种基于跨层次聚合网络的实时城市街景语义分割算法。该算法通过结合跨层次聚合的金字塔池化模块、通道注意力机制以及多尺度融合模块，实现了对多尺度上下文信息的有效提取和特征复用，促进了深层特征与浅层特征的融合。经过在Cityscapes和CamVid两个通用城市街景数据集上的验证，该算法在准确性和实时性之间取得了有效平衡，相较于其他算法在语义分割性能上有显著提升。实验结果表明，该算法在一张RTX3090显卡上以294 FPS的实时性达到73.0% mIoU的准确性，为实时城市街景语义分割领域带来了新的突破。这一研究成果不仅为自动驾驶技术的发展提供了有力支持，也为实时城市街景语义分割领域的研究开辟了新的方向。”摘要：With the rapid development of autonomous driving technology， precise and efficient scene understanding has become increasingly important. Urban street scene semantic segmentation aims to accurately identify and segment elements such as pedestrians， obstacles， roads， and signs， providing necessary road information for autonomous driving technology. However， current semantic segmentation algorithms still face challenges in urban street scene segmentation， mainly manifested in issues such as insufficient discrimination between different categories of pixels， inaccurate understanding of complex scene structures， and inaccurate segmentation of small-scale objects or large-scale structures. To address these issues， this paper proposed a real-time urban street scene semantic segmentation algorithm based on a cross-layer aggregation network. Firstly， a pyramid pooling module combined with cross-layer aggregation was designed at the end of the encoder to efficiently extract multi-scale context information. Secondly， a cross-layer aggregation module was designed between the encoder and decoder， which enhances the representation ability of information by introducing a channel attention mechanism and gradually aggregates the features of the encoder stage to fully achieve feature reuse. Finally， a multi-scale fusion module was designed in the decoder stage， which aggregates global and local information in the channel dimension to promote the fusion of deep and shallow features. The proposed algorithm was validated on two common urban street scene datasets. On an RTX 3090 graphics card （TensorRT speed measurement environment）， the algorithm achieves 73.0% mIoU accuracy on the Cityscapes test set with real-time performance of 294 FPS， and 75.8% mIoU accuracy on higher resolution images with real-time performance of 164 FPS； on the CamVid dataset， it achieves 74.8% mIoU accuracy with real-time performance of 239 FPS. Experimental results show that the proposed algorithm effectively balances accuracy and real-time performance， significantly improving semantic segmentation performance compared to other algorithms， and bringing new breakthroughs to the field of real-time urban street scene semantic segmentation.关键词：semantic segmentation;convolutional neural network;urban street view;encoder-decoder structure;pyramid pooling module
- 49
- |
- 18
- |
- 0
发布时间：2024-05-08 - “钢材缺陷检测领域取得重要突破。针对钢材表面缺陷形态多样、结构复杂等问题，研究者提出了一种轻量化的VTG-YOLOv7-tiny检测算法。该算法通过设计VoVGA-FPN网络，增强特征融合能力；构建三重坐标注意力机制，提升特征提取能力；引入鬼影混洗卷积，降低模型参数量和计算量；增加大目标检测层，改善检测精度。实验验证表明，改进后的算法在NEU-DET和Severstal数据集上mAP分别提升5.7%和8.5%，参数量和计算量分别降低0.61M和4.2G，精确度和召回率也显著提升。这一成果为钢材缺陷检测提供了新的解决方案，并有望为边缘终端设备提供高效、准确的检测能力。”摘要：To address the problems of diverse and complex shapes of steel surface defects， detection target missing， and large number of algorithm parameters， a lightweight VTG-YOLOv7-tiny steel defect detection algorithm was proposed. The method first designed VoVGA-FPN network to reduce the loss of information during information transmission and enhance the network feature fusion ability； second， it constructed a triple coordinate attention mechanism to improve the model's feature extraction ability of spatial and channel information； third， it introduceed ghost shuffle convolution to reduce the model parameters and computation while improving the accuracy； fourth， it added a large target detection layer to improve the problem that some defects in the feature map occupy a large proportion， resulting in low detection accuracy. The improved algorithm was verified on the NEU-DET and Severstal steel defect datasets. Compared with the original model， the mAP of the improved algorithm is increased by 5.7% and 8.5%， respectively； the parameters and computation are reduced by 0.61 M and 4.2 G， respectively； the accuracy and recall are increased by 7.1%， 1.8% and 8.9%， 7.0%， respectively. The experimental results show that the improved algorithm better balances the detection accuracy and lightweight， and provides a reference for edge terminal devices.关键词：defect detection;Lightweight YOLOv7-tiny;VoVGA-FPN network;Triplet Coordinate Attention(TCA);Ghost Shuffle Convolution（GSConv）
- 78
- |
- 72
- |
- 0
发布时间：2024-05-08 - “一项针对工业实际场景的研究取得了重要进展。该研究团队发布了名为PD4CV（Part Detection for Control Valve）2023的密集控制阀零件数据集，为工业生产中的自动目标检测提供了新的资源。该数据集源自控制阀生产车间，包含了9类零件、510张工盘图像和15015个零件样本，具有密集摆放、遮挡、尺寸差异大、外形相似等特点，为自动目标检测带来了诸多挑战。通过对比实验，研究团队发现一般场景数据集和特定工业场景数据集难以应对PD4CV2023数据集的特殊性。然而，一系列目标检测算法在该数据集上的综合对比验证了其有效性，显示出PD4CV2023数据集在一般性目标检测、多尺度目标检测、小规模、不均衡数据下目标检测中的优越性。这一研究成果为面向工业的目标检测研究提供了新的方向，有望推动工业生产中的自动化智能化进程。同时，该数据集也为相关领域的研究人员提供了宝贵的实验资源，为解决工业自动化中的目标检测问题奠定了坚实的基础。”摘要：Automated intelligence in industrial production is inseparable from automatic object detection， and high-accuracy automatic object detection relies on datasets adapted to the actual scene. This article published a dense control valve parts dataset for industrial practical scenarios， named PD4CV （Part Detection for Control Valve） 2023. The image of this dataset came from the control valve production workshop， and after the image collection was completed， it underwent steps such as dataset preprocessing， dataset annotation， and dataset partitioning. The images of this dataset were all from the control valve production workshop. After the image collection was completed， the dataset images were first preprocessed， followed by labeling the part targets in the dataset images. Then， the dataset images were divided into training， validation， and testing sets. The PD4CV2023 dataset covered a total of 9 types of parts， including 510 workstation images and 15 015 part samples， with an average of approximately 29 part samples per image. Compared with the existing object detection datasets， this dataset had the characteristics of dense placement and occlusion of parts， large size differences of parts， similar shapes of some parts， and unbalanced number of parts samples. Finally， pre training comparative experiments on different types of datasets show that general scenario datasets and specific industrial scenario datasets are only suitable for general and specific tasks， while the PD4CV2023 dataset， which represents the actual production conditions of control valves， can be used for target detection of control valve parts， and has its particularity and irreplaceability； a comprehensive comparison of a series of algorithms on this dataset verifies the effectiveness of PD4CV2023 dataset in general object detection， multi-scale object detection， and object detection under small-scale and imbalanced data. The PD4CV2023 dataset can be used for research on industrial oriented object detection algorithms.关键词：deep learning;industrial object detection;dataset;control valve parts
- 61
- |
- 19
- |
- 0
发布时间：2024-05-08 - “在三维重建领域，为了解决不连续位置对基于相位测量轮廓术精度的影响，研究者提出了一种创新的自适应分割方法。该方法通过自适应方向相干性算法，有效确定不连续位置，将相对相位展开为绝对相位。仿真和实际实验均验证了该方法的有效性，显著提高了三维重建的精度。这一研究成果不仅为相位测量轮廓术的发展提供了新的思路，也为高精度三维重建领域的研究开辟了新的方向。”摘要：To mitigate the impact of discontinuous positions on the accuracy of three-dimensional reconstruction using phase measurement profilometry， we introduced an adaptive segmentation method. This method was based on phase measurement and employed a phase discontinuity segmentation algorithm utilizing adaptive orientation coherence.Initially， we employed the adaptive orientation coherence map and the smooth phase map to identify discontinuous positions， that was， the grayscale map of the discontinuous area and the continuous area was obtained. Subsequently， the grayscale map was processed into a binary mask， representing discontinuous areas as 0 and continuous areas as 1. Lastly， this mask map served as a weight map in the weighted least squares phase unwrapping method to convert the relative phase into an absolute phase. The resulting weight primarily assessed the phase quality at each pixel position within the wrapped phase map.The simulation results show that the segmentation mean errors of the adaptive orientation coherence method for discontinuities of straight lines and rectangles in the wrapped phase map with noise variance of 0.8 are 1.678 3 and 3.000 2 pixels， respectively.The actual experiment also proves that the proposed method can effectively segment the discontinuities for high-precision 3D reconstruction.关键词：machine vision;weighted least squares;discontinuity segmentation;mask map;adaptive post-processing
- 65
- |
- 16
- |
- 0
发布时间：2024-05-08
Information Sciences
- Address：No.3888 Dong Nanhu Road, Changchun, Jilin, China Postal code：130033
- Tel：0431-86176855 Email：gxjmgc@ciomp.ac.cn
- Technical support is provided by Beijing Founder electronics co., LTD 吉ICP备11002662号-17
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
0