Search

Accurate detection of infrared small target in complex scenes plays a vital role in many practical applications. In the infrared patch-tensor (IPT) model, small target detection is regarded as a convex optimization problem of tensor robust principal component analysis (TRPCA), which separates low tubal rank and sparse tensors. Tensor singular value thresholding is employd as the proximal operation of tensor nuclear norm to estimate the low-rank background components. Considering the defect that the local structural weights in the Reweighted IPT shrink excessively small targets, an effective local prior map is used to better mine target and background information. The local prior and sparse enhancement weight are combined to obtain the final adaptive weight. Alternating direction method of multipliers (ADMM) is utilized to solve the model. Extensive experiments of infrared sequence images based on ground background prove that our algorithm has significant advantages in real-time and detection capability compared with state-of-the-arts.

The weight of motor is related to torque in exoskeleton. Designers need to install the motor on the exoskeleton. The weight and moment of inertia of the exoskeleton will change when the motor type is changed. For the purpose of determine the motor type, this paper developed a MATLAB Graphical User Interface (GUI) to complete forward kinematics solution, dynamic analysis through program writing and data interaction. The forward kinematics is solved by Denavit-Hartenberg (DH) parameters method, the dynamics is solved by Lagrange equation. General kinematics can be decomposed into multiple polynomials by Taylor series. Input polynomial coefficients into GUI to obtain motor torque curve data. The designer can then determine the motor power by the product of torque and rotational speed.

Being a promising candidate technology for the fifth-generation cellular systems, the ultra-dense network (UDN) is capable of improving the system throughput by deploying more small cells (SCs) in the existing macro-cells (MCs). In most cases, base stations (BSs) from both the MCs tier and SCs tier are modeled as two independent homogeneous Poisson point processes (PPPs) or other point processes derived from a PPP. In the previous works, the area spectral efficiency (ASE) and energy efficiency (EE) are evaluated from a typical user's perspective, and this method works well in single-user scenarios. However, there are multiple users sharing spectral resources in practical cellular systems, and the above-mentioned method cannot work in these scenarios. Therefore, we investigate the ASE and EE from the perspective of a typical BS. Analytical results show that the ASE increases with the increasing SCs density, which means that deploying more SCs in the MCs coverage area is a feasible way to enhance the system throughput. However, the energy consumption of the entire network is increasing as the SCs density increases. Therefore, we employ the firefly algorithm to jointly optimize the ASE and EE for the two-tier heterogeneous UDN. Finally, the optimal system parameters can be found under two fixed weights for the ASE and EE, respectively.

Aiming at the defects that the traditional Census algorithm uses a fixed window and a fixed threshold to cause the image to have discontinuous depths and low matching accuracy in weak texture regions, an improvement is proposed. The cost computation phase uses SAD-Census algorithm, and proposes a new type of adaptive window method. The gradient information is used to dynamically select the threshold value to realize the selection of the window, and the Census cost computation is optimized. Consider the whole picture, Complete cost aggregation at multiple scales based on minimum spanning tree(MST); introduce left and right consistency detection methods to detect mismatched points in occluded areas, smooth the image through singular point filling and median filtering, and improve the overall accuracy of the improved algorithm. Using Middlebury dataset for testing, the experimental results show that the improved algorithm proposed in this paper has significantly improved matching accuracy and robustness compared with traditional algorithms, especially in areas with deep discontinuities and weak textures.

Due to machining errors and assembly errors, the optical system of the camera is somewhat different from the ideal model through the pinholes. That means, there are different degrees of non-linear optical distortion between the ideal image and the actual image formed by the camera. In order to improve the positioning accuracy of the camera, distortion correction should be done. This paper puts forward a GPU-Based Star Calibration and Proofreading Algorithm (GBSC-PA). This algorithm is completed by CUDA programmable graphics processor technologies, and a number of lens distortion coefficients are solved using star point images. Moreover, the optimal distortion correction matrix can be used to improve the accuracy and speed of the algorithm greatly. Through the experimental measurements and simulation, the feasibility and efficiency of the new algorithm are verified. Compared with the conventional correction algorithm based on CPU, the speed-up ratio of GPU-based parallel program can be reached to 43.094.

The use of robots has been proposed as promising human assistants in accomplishing tasks that are otherwise too tough or dangerous to be executed by human alone. As robots in different forms have their differentiating strengths, a system that uses a combination of these robots can be more robust and resilient in achieving its mission. For example, UAVs can be used to quickly explore an unknown environment while ground robots can be used to navigate through the environment, clearing obstacles and potential dangers for human to enter. In this work, we design a unifying framework to allow human to control and accomplish an exploration mission using heterogeneous robot types. A prototype system is implemented using the Google Tango to control an UAV and a wheeled robot for the human to perform environment mapping. Preliminary experiments have been carried out to study the feasibility of the system.

Adding acid control of the circulating water in thermal power plants is one of the important measures of water quality regulation. It is a very complex industry control process, which has the characteristics including nonlinear with description difficulty, pure lag(delay), large inertia, time-varying, etc. For the adding acid control system in thermal power plants, normal PID control can not achieve ideal control effect. On this basis, a new fuzzy PID control strategy based on expert system is put forward in this paper. The simulation experiment is carried out with MATLAB. The results demonstrate that in adding acid control system of circulating water with hysteretic and nonlinear characteristics, the expert fuzzy PID control proposed in this paper can greatly improve the dynamic performance of system which has strong adaptability and robustness, and better control effect is obtained as a result.

The conflicts between the miniaturization of optical remote sensor and the high resolution force us to build an end-to-end simulation system, and optimize the process of the image acquisition and processing. In this paper, based on the analysis of MTF models of each link in imaging chain, a scientific visualization method of end-to-end optical remote sensing process using MTF models is presented. And, imaging simulation software is programmed with Matlab to implement the process. Using this program, the final simulated image can be generated from the original through the whole remote sensing image course according to the input parameters of each link in image chain. Through research on simulated images, we can analyze effectively the effects of each link on the quality of images, and carry out the optimization of optical remote sensor.

Scale invariant feature transform (SIFT) is a better corner extraction algorithm, but there are still mismatching problems in the feature matching step. a new matching principle based on vector similarity is proposed and then it is compared with traditional matching principle. Firstly, the matching feature points are detected by the new principle. Mismatching points are further removed by using the mutual mapping theory. Secondly, transformation matrix is calculated by random sample consensus (RANSAC). Furthermore, the matrix is optimized by Levenberg-Marquardt algorithm (L-M). Lastly, image mosaic is realized by image fusion. Experimental results indicate that compared with traditional matching principle, new matching principle has improved matching accuracy. It is able to apply new principle to image registration and image mosaic.

Traditional LED (light emitting diode) packaging always adopts hemisphere lens, and can ensure high light output efficiency, but its radiation pattern yields Lambert distribution and is non-uniformity. To enable LED light sources be applied in uniform illumination system, a method to obtain freeform LED lens based on energy equipartition and tangent iteration is proposed. Firstly, according to the Lambertian distribution of LED light intensity and its rotationally symmetric peculiarity, the LED emission angle is divided into N parts in Two-dimensional coordinate system, and the energy contained in each part is equal; Secondly, the target plane is divided into N equal areas along its radii, and the discrete points are corresponding to the N parts of emission angle. So that if the energy in each divided solid angle is controlled to illuminate the equal divided area, the illumination of each corresponding micro-belt on the target plane will be uniform; Finally, to avoid the total reflection of large emission angle rays, the designed lens will refract the rays with 0∼450 emission angle and reflect the rays with 450∼900 emission angle. Choosing an initial point on the freeform generatrix, and using Snell equation and tangent iteration method, the coordinates of all points on the freeform generatrix are obtained, which could be rotated into the freeform surface of the lens. The simulation results using the designed LED lens show that for a rotationally symmetric target plane the illumination uniformity is near to 90% and the energy efficiency is higher than 95%.

The number of digital photos in the personal computer is exploding. In an effective photo management system, photo annotation is the most challenging task. The current photo annotation and management systems suffer from two crucial problems. One is the expression of semantic knowledge; the other is the way of photo annotation. Aiming at the former problem, this paper proposes to utilize ontology to organize the domain knowledge and provide formal, explicit and conceptual annotation. Meanwhile, a dual-level semi-automatic annotation approach is also proposed to resolve the latter problem. The rough annotation layer provides preliminary annotation by automatically extracting some semantic concepts from photo titles/texts, time concepts from EXIF metadata, and photo classification concepts from the result of face detection algorithms. The accurate annotation layer provides more detailed annotation by allowing users to modify, delete and add the annotation information freely. An ontology based photo management system OntoAlbum is implemented in this paper. Experimental results show that the proposed approach is very effective and promising.

Highlight detection is a challenge task in soccer video analysis. Using Web-casting text as external knowledge is proved to be a short cut to achieve both efficiency and effectiveness. Based on the previous framework using Web-casting text, we have improved the processes of video time detection and highlight boundary detection. Our method can detect the transparent time bar and can achieve acceptable precision in highlight boundary detection though the Web text time is not accurate at all. This progress can make the framework more robust in practice.