To achieve rapid distance estimation and tracking of moving targets in a large field view, this paper proposes an innovative simulation method. Using low-cost approach, the imaging measurement performance designed cooling-type mid-wave infrared compound-eye camera (CM-CECam) is experimentally evaluated. The consists small-lens array with spherical shell, relay optical system, detector. Based on spatial arrangement array, precise model for developed, constructing virtual space. Distance error...

In bridge buildings, concrete is widely used because its materials are considerably low-cost and it has high plasticity. However, some drawbacks exist in this kind of bridges, crack the most common ones. order to avoid cracks buildings becoming worse, necessary periodically perform inspection for it. Thus, a robot system with machine vision designed precise robust detection. facilitate analysis cracks, number images collected stitched into quality panorama, then crack-like defects panorama...

Web service discovery will become more and important as the services published in Internet expand exponentially. To find a service, non-functional (QoS) properties of services, such time, cost, reliability, play significant role. Though there are many researches work on QoS based matchmaking, few works consider client interests properties. In fact, it is an factor. this paper, we, firstly, propose novel model by adding interest degree property. The computation analyzed, too. Then,...

When deep learning methods are used to detect rail surface defects, the training accuracy declines due small defects and an insufficient number of samples. This paper investigates problem defect detection by using improved feature pyramid network (FPN) metric approach. Firstly, FPN is adding deformable convolution convolutional block attention modules improve detecting different scales, it pretrained on MS COCO dataset. Secondly, a new model established extract features based transfer...

Famous feature point descriptors such as SIFT and SURF allow reliable real-time matching but at a computational cost that limits the number of points can be handled on PCs, even more less powerful mobile devices. A recently proposed technique called Binary Robust Independent Elementary Features (BRIEF) uses binary string an efficient descriptor. It has been shown it is highly discriminative when using relatively few bits. However, we note resulting features are not robust intensity...

Segmentation of the whole-cardiac CT image sequence is key to computer-aided diagnosis and study lesions in heart. Due dilation, contraction flow blood, cardiac images are prone weak boundaries artifacts. Traditional manual segmentation methods time-consuming labor-intensive produce over-segmentation. Therefore, an automatic technique proposed. This was employed using deep learning algorithm understand function from ground truth data. Using convolution neural network (CNN) on central...

For improving the accuracy of traditional maximum entropy threshold segmentation algorithm, an improved algorithm is proposed. Firstly, it determines possible range optimal according to a simple statistical method, so as reduce interference background and magnify proportion target region. Secondly, in certain threshold, does image which obtained by using principle. Simulation experiments show that not only can improve noise immunity effectively, but also better keep details region comparison...

This paper aims to deal with the automatic model composition issue by jointly using service-oriented technology and agent technology. Most existing methods focused on reusability remain static unable self-adapt dynamic environments. The service (SA) concept was proposed enable of component models. main idea solution in this is encapsulate a into so that based can be applied problem. SA method proves perform better environments than traditional one. Finally, case study given system prototype...

Surface defects affect the quality and safety of oil seals. It is a challenge to detect such in vision system because unequal reflection property seal surfaces low contrast between defect background. This article proposes visual detection method (VDM) for surface outlines two key issues VDMs. First, we present superpixel segmentation algorithm based on significant gray level variation radial direction an image. image then divided into several ring belts. Subsequently, considering inequality...

Adoption of machine vision inspection and computer image processing technology, an oil-seal dimension measuring system was developed to meet the requirement online production real time inspection. The makeup principle were introduced, as well its working process design requirements described on detailed. technique quadratic filtering for preprocessing combined with three points determining a circle, point Hough transform least squares employed algorithm, high precision sub-pixel edge...

Lane detection is to determine the precise location and shape of lanes on road. Despite efforts made by current methods, it remains a challenging task due complexity real-world scenarios. Existing approaches, whether proposal-based or keypoint-based, suffer from depicting effectively efficiently. Proposal-based methods detect distinguishing regressing collection proposals in streamlined top-down way, yet lack sufficient flexibility lane representation. Keypoint-based other hand, construct...

A robot mapping algorithm based on the gray system theory is described in this paper. The ultrasonic error model established according to sound wave transmission character and a value defined express sensor data uncertainty. In algorithm, several continuous sonar are fused fusion order update of map grid. grid neighborhood searching method proposed judge accessible position, plan feasible path for complete whole environment. validity accuracy proved real office

This paper proposes an algorithm of building keypoint matches on multimodal images by combining a bootstrap process and global information. The correct ratio built with descriptors is typically very low large spectral difference. To identify matches, information utilized for evaluating technique employed to reduce the computational cost. A match determines transformation <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math...

<title>Abstract</title> In power line inspections, detecting insulator defects is critical due to the potential breakdown and damage resulting from long-term exposure natural environment. However, challenges persist in practical detection processes, such as scarce defect images, difficulty minor defects, high memory consumption of deep learning models. We propose an algorithm based on adaptive feature fusion lightweight YOLOv5s model address these issues. First, we enhance existing dataset...

This paper presents a method to recognize continuous full-body human motion online by using sparse, low-cost sensors. The only input signals needed are linear accelerations without any rotation information, which provided four Wiimote sensors attached the limbs. Based on fused hidden Markov model (FHMM) and autoregressive process, predictive fusion (PFM) is put forward, considers different influences of upper lower limbs, establishes HMM for each part, fuses them probabilistic model. Then an...