CPA Algorithm for S, R circle parameters in the determination of circular building objects using RTK/NRTK GNSS technology and linear measurements

A fast Hough transform for circular object detection is proposed in this paper which can be directly applied to gray level images. This method consists of three major stages. In the first stage, the center positions of circular objects are detected using the gray level Hough transform, which requires no conventional preprocessing such as edge detecting and binarization. The second stage determines the radius of the detected objects by analyzing the radial gradient profile. In order to detect objects with different radius in the same scene, a multi-scale strategy is integrated in the proposed method. Compared with traditional Hough transform, the gray level Hough transform uses a 2-dimensional accumulation map rather than the 3-dimensional one, which results in a dramatic improvement on the computational efficiency. Experiments have been carried out on more than 2000 real-world images and the result shows that 90.3% of the circular objects have been accurately detected, which demonstrate the applicability of the proposed method.

Vision guided robots have more ability, functionality and adaptivity in industrial assembly lines than normal robots. This research attempts to increase the impact of computer vision on robotic positioning and grasping applications. Therefore, we addressed object detection and localization to perform robotic grasping and positioning using Selective Compliant Assembly Robot Arm (SCARA). The target position of SCARA robot is determined based on information obtained from object detection and position measurement process. This process is implemented on a circular object to simplify the task. For accurate position measurement, the distortion of camera lens is removed using camera calibration technique. In object detection, several methods are compared to detect circular holes in an input image. The most successful methods with 100% Precision, Recall and F-measure are used to detect the circular object. The position of this object is measured in world coordinate unit for pick-and-place operation. Then, the experiment is designed to move SCARA robot to the measured position of the detected circular object. The result showed that the robot is successfully moved to the measured position of the detected object with average positioning error (0.314, 0.155) mm.

Real-time moving object tracking using radar is a technology in high demand because of the penetration properties of radar signals and the advantage in the aspect of security, when compared to conventional optical systems. In this research, an ultra-wideband impulse signal was used to achieve high resolution in inverse synthetic aperture radar images. It is possible to track the position of moving objects in real-time by using one transmitting antenna and multi receiving antenna.

Object detection is a critical step of many image recognition systems. In this paper, we discussed the circular shape object detection problem in still color images. The proposed method has an important feature of integrating color image edge extraction result with a novel circle parameter determination algorithm efficiently. First, an accurate isotropic edge detector is introduced to extract color image edge and calculate precise gradient direction of each potential edge pixel, which assures the high accuracy of subsequent circle detection. Then the detected potential edge results are verified by integrating them with spatial information and the result of region based analysis. Second we utilize just only one 2-dimensional accumulator array and one 1-dimensional accumulator array, which greatly reduce storage requirement and time complexity, to detect circles of any radius. Furthermore, the validity of detected potential circle centers is checked to avoid detecting false circles. Experimental results show that our method is robust and effective in locating objects with complete or incomplete or concentric circle boundary in real color images without any prior knowledge.

Ultrasound imaging offers a safe, non-invasive, non-ionizing tool for clinical diagnosis. Ultrasound tomography would allow ultrasound imaging to reach its full potential, compared to X-ray, MRI, etc. It is based on inverse scattering and has the capability to resolve structures which are smaller than the wavelength of the incident wave, as opposed to conventional ultrasound imaging using echo method. Some material properties such as sound contrast are very useful to detect small objects. The distorted Born iterative method based on first-order Born approximation has been introduced as an efficient diffraction tomography approach. In this paper, we propose to replace linear measurements by random ones in the detection geometry configuration. Reconstructions of circular cylindrical objects including 4.5% of speed of sound in a grid size of 32 × 32 for full 360 degree arrangement of transmitters and detectors are presented. The quality of reconstruction is improved and the complexity is reduced.

In recent years, 3D object detection has attracted considerable interest from researchers due to its crucial importance in scene understanding tasks for mobile robots. The task of 3D object detection is to determine the position, orientation, scale, and category of objects in a scene, with multimodal data inputs. These inputs consist of two distinct types: images and point clouds. Images provide colour information about the scene, while point clouds deliver geometric information. The fusion of these two diverse data types can enhance the detection of 3D objects by allowing them to complement each other. In this paper, a review of real-time multimodal 3D object detection from the perspective of multimodal sensor fusion is presented. It is important to note that previous reviews of multimodal 3D object detection have been limited to autonomous driving scenarios. In contrast, this paper reviews multimodal 3D object detection approaches in both indoor and outdoor scenarios. In addition, the popular datasets and assessment metrics employed for 3D object detection are described, as well as the issues and future prospects of 3D object detection.

The issue on an aged deterioration to a reinforced concrete (RC) building is revealing in recent years. Especially, a falling cover concrete is an urgent subject since it may cause harm to the surrounding people. Number of experimental studies on an aged deterioration have been performed to relate the corrosion cracking appearance on the concrete surface with the corrosion ratio of steel bars. It seems that corrosion cracking propagation also induces cover concrete to peel off. However, there is almost no study on the risk of exfoliation of a cover concrete. Accelerated corrosion tests were conducted to estimate the splitting tensile strength before and after corrosion cracking appearance on the concrete surface and to assess the risk of falling cover concrete. The specimens are concrete cylinders having 150 mm in diameter and 300 mm in length, and a single reinforcement of D16 was installed right in the center of the cylinder. All surfaces except for exposed opposite faces of 50 mm in width were coated by epoxy resin to avoid penetration of NaCl solution. The variable factor is accumulative current density (0, 30, 100, 200 and 400 mA·h/cm2). The specimens were cured for 28 days before accelerated corrosion test. During the accelerated corrosion test, the specimens were placed in the tank containing 3 % of NaCl solution. A direct-constant 10 voltage was given and the current was monitored and recorded using a data logger in 60 minutes increments. The cracks on the concrete surface were visually observed and the crack width at certain locations was frequently measured using a digital microscope having a resolution of 0.01 mm. After accelerated corrosion tests, the corroded steel bars were removed from the surrounding concrete by performing a split tensile test to evaluate a splitting tensile strength on the corrosion cracking surface. A removed bar was soaked into 10% diammonium hydrogen citrate solution. After chemically and mechanically cleaning, the weight loss of corroded rebars was measured. Since a corrosion rate before cracking was slower and increased after cracking, a corrosion ratios was estimated from the accumulative current density and using a bi-linear interpolation having a break point at cracking. The circular cross section was smaller than the rectangular cross section, and the pores to accumulate a corrosion product decreased, as a result, the corrosion crack width of a circular specimen became larger than that of a rectangular specimen. Split tensile test results indicated that the splitting tensile strength before cracking (up to corrosion ratio of 1.7%) nearly didn't deteriorate due to an occasional stain of corrosion product on the concrete surface, however, it gradually decreased after outflowing of corrosion product from a connected crack. It was reduced by 26% and 37% when the corrosion ratio increased to 4% and 10% respectively. A simple model was proposed to assess a retention factor of splitting tensile strength with increasing corrosion. Furthermore, a finite element analysis was conducted on the split tensile specimen to investigate a fracture energy of a cracked surface by corrosion because the corrosion crack did not progress uniformly but was unevenly distributed to the splitting surface. Analytical results showed that the mechanical behavior of the splitting tension test was overwhelmingly dominated by the tensile softening properties of the splitting surface and not only the splitting tensile strength but the crack width should be reduced similarly to obtain a result to be consistent with an experimental strength. Further study is required to improve a prediction of falling concrete from a width of corrosion cracking on the concrete surface due to the sensitivity to tensile softening properties and corrosion level.

The edge effect is a measurement error resulting from the reflection of the laser beam on the ad- jacent walls, or by its breaking on the edges. Coordinates of points in such cases are determined by averaging the measurements in several areas, resulting in their incorrect positioning in space. Corner points are determined with the same accuracy as the other (flat) elements of the scanned object. This effect is frequently mentioned in publications, which typically state the reasons and mechanisms of the error thus occurred. However, there is a lack of specific examples, showing the impact of the edge effect on the quality and accuracy of geodetic and cartographic reports. In this paper, the authors present sample case studies of the 2D and 3D representation of the test object. The selected corner elements, as well as the vector elements fitted into a cloud of points, show the discrepancy between the breaking points in the drawing (model), and the curved sur- face of the point cloud. On the basis of the known geometry of the building, distances were de- termined between the corner points and their representatives on the cloud. In this way, we were able to determine the accuracy of corner points’ presentation by means of the cloud of points, and therefore, we were able to determine the size of the edge effect in specific cases.

Discovering and restoring changes in object positions using an autonomous robot with laser rangefinders

Video object segmentation (VOS) is an essential part of intelligent transportation systems (ITS). In the real traffic scenes, not only stable accuracy but also real-time speed are important metrics. In this paper, we propose a semi-supervised real-time video object segmentation method based on a Siamese network with a novel polar representation method. This polar representation method could reduce the parameters for encoding masks, so that the inference speed is enhanced significantly. Besides, an asymmetric Siamese network is designed to extract the features from different spatial scales. To reduce the antagonism among the branches of the head part, we propose the idea of peeling convolution. Based on this idea, repeated cross-correlation and semi-FPN are designed as the neck part of the neural network. The experiments on the DAVIS-2016 dataset demonstrate that SiamPolar achieves 71.4% J-mean and 59.2fps. In the real traffic scene dataset, TSD-max dataset, SiamPolar performs the 80.5% J-mean.

Owing to the development of graphical user interfaces, computers have become a common tool for use by all age groups both at home and at work. Furthermore, touchscreen-based user interfaces are now commonly used in mobile computers such as smartphones and tablets. When we anticipate that future computers will become invisible, that is, computing functionality will be completely embedded into household appliances, furniture, and houses, our interactions will not be limited to things on a display or a tabletop. In such situations, pleasantness may become important as well as accuracy and efficiency in interface design. This study presents a new scheme for detecting 3D positions of objects in water, based on our experiences with the development of aquatic interactive systems. For the detection of objects on a certain 2D layer in water, a line laser is used with a camera. Stacking multiple line lasers enables the detection of object positions in a 3D space.

In this paper, the stereo vision technique with two cameras has been employed to locate object's position in 3D space. It is necessary to use many algorithms to derive object positions, including stereo calibration, rectification, correspondence and triangulation. Two tracking algorithms have been applied to identify and track objects alternately: template matching and Camshift. To evaluate the proposed approach, the stereo vision system is installed on a cubic space which is the working space of a tiny robot arm. The real positions in this coordinate system are then compared to the received positions from the cameras. The end-effector of the robot arm will track the object depending on received positions. Experiment results show that the robot arm can effectively track the indicated objects, and thus the tracking errors between actual and received positions are less than 3% which are reasonably acceptable.

Three-dimensional planar object tracking with sub-pixel accuracy

Changes in visual perception at the time of saccades

The interest in fuzzy algorithms is increasing in a wide range of pattern recognition applications. This paper describes a recognition system using fuzzy algorithms and parameters, with particular enhancement in the system architecture. This system is able to recognize flat polygonal objects in real time. The main aim of the system is to perform an object recognition based on contour description and being independent of noise and indeterminations in the digitization process. This system is made up of three modules. The first one is a break points detector. The second block performs a contour description by means of fuzzy parameters that allows us to recognize the picture in the third block. >