Target Edge Research Articles

Research on calculation strategy of perimeter of irregular objects

Image processing is a burgeoning subject in recent decades, which combines related theories with modern high technology to systematically study various image theories, technologies and applications. In the problem of industrial precision dimension detection, the edge of the obtained target image is rough and noisy due to the manufacturing accuracy, breakage and corrosion of the workpiece, which makes automatic detection difficult. Therefore, it has become a difficult problem to accurately extract the characteristic curve of the target edge in the development of industrial automatic detection system. Image measurement method has the advantages of non-contact, high speed, large dynamic range and rich information, and is widely used in the field of measurement. This paper introduces the principle of foot type three-dimensional contour measurement based on the light cutting method. Aiming at the composition of a multi CCD light cutting measurement system, a new method for high-precision extraction of the center line of the contour light band using the combination of central axis tracking and improved deflection measurement principle is proposed. The extraction and processing process of the light cutting surface contour image is described in detail, and through the coordinate transformation of the extracted center line, The complete profile data of the foot profile are obtained. The experimental results show that the processing effect of this method is satisfactory.

AFSFusion: An Adjacent Feature Shuffle Combination Network for Infrared and Visible Image Fusion

To obtain fused images with excellent contrast, distinct target edges, and well-preserved details, we propose an adaptive image fusion network called the adjacent feature shuffle-fusion network (AFSFusion). The proposed network adopts a UNet-like architecture and incorporates key refinements to enhance network architecture and loss functions. Regarding the network architecture, the proposed two-branch adjacent feature fusion module, called AFSF, expands the number of channels to fuse the feature channels of several adjacent convolutional layers in the first half of the AFSFusion, enhancing its ability to extract, transmit, and modulate feature information. We replace the original rectified linear unit (ReLU) with leaky ReLU to alleviate the problem of gradient disappearance and add a channel shuffling operation at the end of AFSF to facilitate information interaction capability between features. Concerning loss functions, we propose an adaptive weight adjustment (AWA) strategy to assign weight values to the corresponding pixels of the infrared (IR) and visible images in the fused images, according to the VGG16 gradient feature response of the IR and visible images. This strategy efficiently handles different scene contents. After normalization, the weight values are used as weighting coefficients for the two sets of images. The weighting coefficients are applied to three loss items simultaneously: mean square error (MSE), structural similarity (SSIM), and total variation (TV), resulting in clearer objects and richer texture detail in the fused images. We conducted a series of experiments on several benchmark databases, and the results demonstrate the effectiveness of the proposed network architecture and the superiority of the proposed network compared to other state-of-the-art fusion methods. It ranks first in several objective metrics, showing the best performance and exhibiting sharper and richer edges of specific targets, which is more in line with human visual perception. The remarkable enhancement in performance is ascribed to the proposed AFSF module and AWA strategy, enabling balanced feature extraction, fusion, and modulation of image features throughout the process.

A Belief-Based Task Offloading Algorithm in Vehicular Edge Computing

In vehicular edge computing (VEC), where vehicles offload their tasks to nearby edge clouds, it is not a trivial issue to design an optimal task offloading policy due to the dynamic nature of VEC environment and limited information on computing and communication resources. In this paper, we propose a belief-based task offloading algorithm (BTOA) where a vehicle selects target edge clouds (for computing) and subchannels (for communications) based on its belief, and observe their current resource and channel conditions. Based on the observed information, the vehicle finally determines the most appropriate edge cloud and subchannel. Evaluation results under a realistic traffic scenario demonstrate that BTOA can reduce the total latency of the task offloading over <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$42\%$</tex-math> </inline-formula> compared to a conventional offloading algorithm where the target edge clouds and subchannels are determined without any real observations.

Small Target Detection Method Based on Low-Rank Sparse Matrix Factorization for Side-Scan Sonar Images

Target detection in side-scan sonar images plays a significant role in ocean engineering. However, the target images are usually severely interfered by the complex background and strong environmental noise, which makes it difficult to extract robust features from small targets and makes the target detection task quite challenging. In this paper, a novel small target detection method in sonar images is proposed based on the low-rank sparse matrix factorization. Initially, the side-scan sonar images are preprocessed so as to highlight the individual differences of the target. Then, the problems of target feature extraction and noise removal are characterized as the problem of matrix decomposition. An improved Robust Principal Component Analysis algorithm is used to extract target information, and the fast proximal gradient method is used to optimize the solution. The original sonar image is reconstructed into the low-rank background matrix, the sparse target matrix, and the noise matrix. Eventually, a morphological operation is used to filter out the noise and refine the target edges in the target matrix for improving the accuracy of target detection. Experimental results show that the proposed method not only achieves better detection performance in comparison to the conventional baseline algorithms but also performs robustly in various signal-to-clutter ratio conditions.

Deep Learning-Based Semantic Segmentation Methods for Pavement Cracks

As road mileage continues to expand, the number of disasters caused by expanding pavement cracks is increasing. Two main methods, image processing and deep learning, are used to detect these cracks to improve the efficiency and quality of pavement crack segmentation. The classical segmentation network, UNet, has a poor ability to extract target edge information and small target segmentation, and is susceptible to the influence of distracting objects in the environment, thus failing to better segment the tiny cracks on the pavement. To resolve this problem, we propose a U-shaped network, ALP-UNet, which adds an attention module to each encoding layer. In the decoding phase, we incorporated the Laplacian pyramid to make the feature map contain more boundary information. We also propose adding a PAN auxiliary head to provide an additional loss for the backbone to improve the overall network segmentation effect. The experimental results show that the proposed method can effectively reduce the interference of other factors on the pavement and effectively improve the mIou and mPA values compared to the previous methods.

Research on Driving Obstacle Detection Technology in Foggy Weather Based on GCANet and Feature Fusion Training.

The issues of the degradation of the visual sensor's image quality in foggy weather and the loss of information after defogging have brought great challenges to obstacle detection during autonomous driving. Therefore, this paper proposes a method for detecting driving obstacles in foggy weather. The driving obstacle detection in foggy weather was realized by combining the GCANet defogging algorithm with the detection algorithm-based edge and convolution feature fusion training, with a full consideration of the reasonable matching between the defogging algorithm and the detection algorithm on the basis of the characteristics of obvious target edge features after GCANet defogging. Based on the YOLOv5 network, the obstacle detection model is trained using clear day images and corresponding edge feature images to realize the fusion of edge features and convolution features, and to detect driving obstacles in a foggy traffic environment. Compared with the conventional training method, the method improves the mAP by 12% and recall by 9%. In contrast to conventional detection methods, this method can better identify the image edge information after defogging, which significantly enhances detection accuracy while ensuring time efficiency. This is of great practical significance for improving the safe perception of driving obstacles under adverse weather conditions, ensuring the safety of autonomous driving.

Auxiliary-Task-Based Energy-Efficient Resource Orchestration in Mobile Edge Computing

Advances in edge computing significantly impact the development of mobile networks. As the most important research goal related to edge networks, resource orchestration has been well studied in recent years; however, existing approaches based on deep reinforcement learning share similar bottlenecks in training inefficiency. In this paper, we treat drones, whose available time is significantly limited by their batteries, as the mobile terminals of a target edge network and aim to maximize the energy efficiency. The battery-constrained resource orchestration problem is formulated as a nonconvex optimization problem with consideration of both operating costs and available battery. Owing to the NP-hard nature of mixed-integer programming, the Auxiliary-Task-based dynamic Weighting Resource Orchestration (ATWRO) algorithm is proposed. To improve the sample efficiency, related parameters serving as auxiliary tasks are employed to provide additional gradient information. We further refine the exploration space and apply an alternative replay buffer to develop a customized reinforcement learning approach. Extensive experiments demonstrate the effectiveness of the proposed scheme, as they prove that by employing auxiliary tasks, reinforcement learning agents can be trained with higher efficiency. Moreover, the service time of the whole system can be prolonged, and a higher number of completed tasks can be guaranteed.

Electron and ion acceleration from femtosecond laser-plasma peeler scheme

Using three-dimensional particle-in-cell simulations, we further investigate the electron and ion acceleration from femtosecond laser–plasma peeler scheme which was proposed in our recent paper (Shen et al 2021 Phys. Rev. X 11 041002). In addition to the standard setup where a laser pulse impinges on an edge of a single tape target, two new variants of the target, i.e. a parallel tape and a cross tape target, were proposed, where strong surface plasma waves can also be efficiently excited at the front edges of the target. By using a tabletop 200 TW-class laser pulse, we observe generation of high-flux, well-collimated, superponderomotive electrons. More importantly, quasimonoenergetic proton beams can always be obtained in all the three setups, while with the single tape case, the obtained proton beam has the highest peak energy and narrowest spectrum.

Sunspot extraction and hemispheric statistics of YNAO sunspot drawings using deep learning

Sunspot drawings around the globe provide long historical records for understanding the long-term trends in the solar activity cycle. Yunnan Astronomical Observatory (YNAO) in China contributes to the relatively continuous sunspot drawings from 1957 to 2015. This paper proposes a new deep learning method named SPR-mask to extract pores, spots, umbrae and penumbrae in the YNAO sunspot drawings. SPR-mask consists of three parts: backbone, shared head and mask branch. It especially adopts a scale-aware attention network (SAAN) and a PointRend module in the mask branch to improve the accuracy of target edge segmentation. Besides that, each sunspot belonging to the northern or southern (N-S) hemisphere is determined by transforming its cartesian coordinates to spherical coordinates after extracting $P$ , $B_{0}$ and $L_{0}$ handwritten in sunspot drawings using a revised Lenet-5 deep learning method. The precision, recall and $AP$ of SPR-mask are 0.92, 0.93, and 0.92, respectively. The test results show the SPR-mask method has a good performance. The numbers and areas of pores, spots, umbrae and penumbrae for the N-S hemisphere are presented and analyzed separately. The YNAO data are also compared with Royal Greenwich Observatory (RGO), Kanzelhöhe Observatory (KSO) and Purple Mountain Astronomical Observatory (PMO) data. The results show similar trends, high correlations, and N-S asymmetries. All data of YNAO are publicly shared at https://github.com/yzs64/YNAO_sd/ , which are abundant and complementary to the other sunspot catalogs in the world.

Which Target to Focus on: Class-Perception for Semantic Segmentation of Remote Sensing

Deep Learning-based (DL) methods have dominated the task of semantic segmentation of remote sensing images. However, the sizes of different objects vary widely, and there is a great deal of label-noise due to the inevitable shadows. Therefore, there is an urgent need for a method that can precisely handle complex ground data. In this paper, we propose an Inter-Class Enhanced Network (ICEN) for representing features of varying sizes. It comprises two branches: Sparse Representation Network (SPN) and Feature Extraction Network (FEN). Then, a Class-Perception Block is inserted between the two branches to instruct the SPN’s low-level semantic features to be merged into the deeper network. Such a block can reduce label-noise in remote sensing image segmentation. In addition, the proposed EIRI provides a more precise classification process for target edges containing many misclassified points without requiring excessive computational overhead. The experimental results of our proposed Class-Perception Network (C-PNet) achieve competitive performance on the Vaihingen, Potsdam, LoveDA, and UAVid datasets.

Algorithm for Intelligent Recognition Low-Grade Seismic Faults Using Codec Target Edges

The development of remaining oil plays an important role in increasing the late production, and low-grade faults seriously impacts residual oil exploration and development. Low-grade faults have a small fault displacement, brief extension, and strong concealment, which hinders their prediction using the typical semantic segmentation network. To intelligently identify low-grade faults, we designed a codec target edge detection technique. For the network to fully learn the low-grade faults information, we constructed the encoder using dilated convolution. Next, we introduced an attention mechanism to the decoder to improve capture of location information from a shallow network and semantic information from a deep network. Finally, the multi-scale fusion decoder outputs fault information of different scales, which further improves the identification accuracy of low-grade faults. The training model is applied to simulated data and actual seismic data through ablation experiments. The results show that this method can effectively identify low-grade faults and overcomes problems such as blurred fault cross-location, thicker edge contour lines, lower detection accuracy, and less training data. Compared with conventional a Holistically-Nested Edge Detection (HED) and a semantic segmentation (UNet), fault misidentification is reduced, fault continuity is increased, and fault accuracy is improved. providing technical support for the exploration and development of remaining oil and increasing the recovery rate of old oil fields.

Service migration for mobile edge computing based on partially observable Markov decision processes

With the continuous development of mobile edge computing, people are more willing to offload tasks to edge servers that are closer to users than cloud services for a better user experience. Due to problems such as user mobility and limited coverage of edge servers, how to ensure service quality and prevent service interruption is a problem we have to consider. This article explores the problem of service migration to decide when, where, and how to migrate an ongoing service from the current edge server to the target edge server. At the same time, we will consider the premise that edge servers can only obtain partial user information, and model the service migration problem as a partially observable Markov decision process. To minimize user delay and system energy consumption, we propose a service migration decision algorithm based on Deep recurrent Q-learning (DRQNSM). Multiple experiments show that our algorithm has a better performance than some classical reinforcement learning algorithms.

Edge-aware object pixel-level representation tracking

Recently, there has been a trend in tracking to use more refined segmentation mask instead of coarse bounding box to represent the target object. Some trackers proposed segmentation branches based on the tracking framework and maintain real-time speed. However, those trackers use a simple FCNs structure and lack of the edge information modeling. This makes performance quite unsatisfactory. In this paper, we propose an edge-aware segmentation network, which uses the complementarity between target information and edge information to provide a more refined representation of the target. Firstly, We use the high-level features of the tracking backbone network and the correlation features of the classification branch of the tracking framework to fuse, and use the target edge and target segmentation mask for simultaneous supervision to obtain an optimized high-level feature with rough edge information and target information. Secondly, we use the optimized high-level features to guide the low-level features of the tracking backbone network to generate more refined edge features. Finally, we use the refined edge features to fuse with the target features of each layer to generate the final mask. Our approach has achieved leading performance on recent pixel-wise object tracking benchmark VOT2020 and segmentation datasets DAVIS2016 and DAVIS2017 while running on 47 fps. Code is available at https://github.com/TJUMMG/EATtracker.

Treatment Planning of Bulky Tumors Using Pencil Beam Scanning Proton GRID Therapy.

To compare spatially fractionated radiation therapy (GRID) treatment planning techniques using proton pencil-beam-scanning (PBS) and photon therapy. PBS and volumetric modulated arc therapy (VMAT) GRID plans were retrospectively generated for 5 patients with bulky tumors. GRID targets were arranged along the long axis of the gross tumor, spaced 2 and 3 cm apart, and treated with a prescription of 18 Gy. PBS plans used 2- to 3-beam multiple-field optimization with robustness evaluation. Dosimetric parameters including peak-to-edge ratio (PEDR), ratio of dose to 90% of the valley to dose to 10% of the peak VPDR(D90/D10), and volume of normal tissue receiving at least 5 Gy (V5) and 10 Gy (V10) were calculated. The peak-to-valley dose ratio (PVDR), VPDR(D90/D10), and organ-at-risk doses were prospectively assessed in 2 patients undergoing PBS-GRID with pretreatment quality assurance computed tomography (QACT) scans. PBS and VMAT GRID plans were generated for 5 patients with bulky tumors. Gross tumor volume values ranged from 826 to 1468 cm3. Peak-to-edge ratio for PBS was higher than for VMAT for both spacing scenarios (2-cm spacing, P = .02; 3-cm spacing, P = .01). VPDR(D90/D10) for PBS was higher than for VMAT (2-cm spacing, P = .004; 3-cm spacing, P = .002). Normal tissue V5 was lower for PBS than for VMAT (2-cm spacing, P = .03; 3-cm spacing, P = .02). Normal tissue mean dose was lower with PBS than with VMAT (2-cm spacing, P = .03; 3-cm spacing, P = .02). Two patients treated using PBS GRID and assessed with pretreatment QACT scans demonstrated robust PVDR, VPDR(D90/D10), and organs-at-risk doses. The PEDR was significantly higher for PBS than VMAT plans, indicating lower target edge dose. Normal tissue mean dose was significantly lower with PBS than VMAT. PBS GRID may result in lower normal tissue dose compared with VMAT plans, allowing for further dose escalation in patients with bulky disease.

Automatic Calibration between Multi-Lines LiDAR and Visible Light Camera Based on Edge Refinement and Virtual Mask Matching

To assist in the implementation of a fine 3D terrain reconstruction of the scene in remote sensing applications, an automatic joint calibration method between light detection and ranging (LiDAR) and visible light camera based on edge points refinement and virtual mask matching is proposed in this paper. The proposed method is used to solve the problem of inaccurate edge estimation of LiDAR with different horizontal angle resolutions and low calibration efficiency. First, we design a novel calibration target, adding four hollow rectangles for fully automatic locating of the calibration target and increasing the number of corner points. Second, an edge refinement strategy based on background point clouds is proposed to estimate the target edge more accurately. Third, a two-step method of automatically matching between the calibration target in 3D point clouds and the 2D image is proposed. Through this method, i.e., locating firstly and then fine processing, corner points can be automatically obtained, which can greatly reduce the manual operation. Finally, a joint optimization equation is established to optimize the camera’s intrinsic and extrinsic parameters of LiDAR and camera. According to our experiments, we prove the accuracy and robustness of the proposed method through projection and data consistency verifications. The accuracy can be improved by at least 15.0% when testing on the comparable traditional methods. The final results verify that our method is applicable to LiDAR with large horizontal angle resolutions.

Deep Multi-Task Learning for an Autoencoder-Regularized Semantic Segmentation of Fundus Retina Images

Automated segmentation of retinal blood vessels is necessary for the diagnosis, monitoring, and treatment planning of the disease. Although current U-shaped structure models have achieved outstanding performance, some challenges still emerge due to the nature of this problem and mainstream models. (1) There does not exist an effective framework to obtain and incorporate features with different spatial and semantic information at multiple levels. (2) The fundus retina images coupled with high-quality blood vessel segmentation are relatively rare. (3) The information on edge regions, which are the most difficult parts to segment, has not received adequate attention. In this work, we propose a novel encoder–decoder architecture based on the multi-task learning paradigm to tackle these challenges. The shared image encoder is regularized by conducting the reconstruction task in the VQ-VAE (Vector Quantized Variational AutoEncoder) module branch to improve the generalization ability. Meanwhile, hierarchical representations are generated and integrated to complement the input image. The edge attention module is designed to make the model capture edge-focused feature representations via deep supervision, focusing on the target edge regions that are most difficult to recognize. Extensive evaluations of three publicly accessible datasets demonstrate that the proposed model outperforms the current state-of-the-art methods.

Hardware Deployable Edge-AI Solution for Prescreening of Oral Tongue Lesions Using TinyML on Embedded Devices

Diagnosing oral cavity cancer (OCC) in its initial stages is an effective way to reduce patient mortality. However, current prescreening solutions are manual and the resultant clinical treatment is not cost effective for the average individual, primarily in developing nations. In this letter, we present an automated and inexpensive prescreening solution utilizing artificial intelligence (AI) deployed on embedded edge devices to detect benign and premalignant superficial oral tongue lesions. The proposed machine vision solution utilizes a clinically annotated photographic dataset of nine types of superficial oral tongue lesions to retrain a MobileNetV2 neural network using transfer learning. In this approach, we also utilized TensorFlow Lite for Microcontrollers to quantize a 32-bit floating point (float32) precision model into an 8-bit integer (int8) model for deployment on power and resource-constrained OpenMV Cam H7 Plus embedded edge device. The quantized int8 model was able to detect the nine tongue lesions with an accuracy of 98.69% on the test set. More than 60% reduction in on-device RAM and flash memory usage were measured for the int8 model when compared to the equivalently performing float32 model for relatively the same inference speed (~1.1 ms) on the target edge device.

An improved anchor-free SAR ship detection algorithm based on brain-inspired attention mechanism.

As a computing platform that can deal with problems independently and adapt to different environments, the brain-inspired function is similar to the human brain, which can effectively make use of visual targets and their surrounding background information to make more efficient and accurate decision results. Currently synthetic aperture radar (SAR) ship target detection has an important role in military and civilian fields, but there are still great challenges in SAR ship target detection due to the problems of large span of ship scales and obvious feature differences. Therefore, this paper proposes an improved anchor-free SAR ship detection algorithm based on brain-inspired attention mechanism, which efficiently focuses on target information ignoring the interference of complex background. First of all, most target detection algorithms are based on the anchor method, which requires a large number of anchors to be defined in advance and has poor generalization capability and performance to be improved in multi-scale ship detection, so this paper adopts an anchor-free detection network to directly enumerate potential target locations to enhance algorithm robustness and improve detection performance. Secondly, in order to improve the SAR ship target feature extraction capability, a dense connection module is proposed for the deep part of the network to promote more adequate deep feature fusion. A visual attention module is proposed for the shallow part of the network to focus on the salient features of the ship target in the local area for the input SAR images and suppress the interference of the surrounding background with similar scattering characteristics. In addition, because the SAR image coherent speckle noise is similar to the edge of the ship target, this paper proposes a novel width height prediction constraint to suppress the noise scattering power effect and improve the SAR ship localization accuracy. Moreover, to prove the effectiveness of this algorithm, experiments are conducted on the SAR ship detection dataset (SSDD) and high resolution SAR images dataset (HRSID). The experimental results show that the proposed algorithm achieves the best detection performance with metrics AP of 68.2% and 62.2% on SSDD and HRSID, respectively.

AMS-PAN: Breast ultrasound image segmentation model combining attention mechanism and multi-scale features

Breast ultrasound medical images are characterized by poor imaging quality and irregular target edges. During the diagnosis process, it is difficult for physicians to segment tumors manually, and the segmentation accuracy required for diagnosis is high, so there is an urgent need for an automated method to improve the segmentation accuracy as a technical tool to assist diagnosis. This study designed an improved Pyramid Attention Network combining Attention mechanism and Multi-Scale features (AMS-PAN) for breast ultrasound image segmentation. On the encoding side, the model adopts the depthwise separable convolution strategy to achieve a multi-scale receptive field with cumulative small-size convolution, which performs multi-dimensional feature extraction and forms a feature pyramid. The model uses Global Attention Upsample (GAU) feature fusion on the decoding side. In order to further process the fused feature information, the proposed method uses a Spatial and Channel Attention (SCA) module to shift the model’s segmentation focus to the edge texture information. The good segmentation performance of our method is verified through experiments on BUSI and OASBUD. All the designed parts have contributed to the segmentation performance in practical applications. Compared with the traditional non-deep learning methods and the current mainstream deep learning methods, the improvement of the model in Dice and IoU metrics is pronounced. AMS-PAN has high computational efficiency, and its good performance has been proven to play a role in ultrasound detection tasks of breast tumors for physicians to specific auxiliary diagnostic roles to guide the subsequent diagnosis and treatment services for patients.

Research on ice characteristics in Bohai Bay, China based on Multi-source remote sensing images

Sea ice area accounts for about 22.9% of the global ocean area, so the research of ice plays an important role in the marine field. The ice data that can be obtained by means of site investigation is very limited, and remote sensing means has unique advantages in large-scale and multi-element feature extraction. Especially in recent years, the available remote sensing image data is becoming more and more abundant, which makes it an important supplementary data for ice research. In this paper, we took the typical sea area of Bohai Bay in China as an example, analyzed the overall characteristics of the ice condition in the target sea area by collecting the winter remote sensing image data in recent 8 years. By using the methods of target recognition, edge detection and time series analysis, four main characteristics of the ice blocks were extracted: time characteristics, spatial characteristics, scale characteristics and motion characteristics. The results enriched the research data of ice condition in the target sea area.