Ground penetrating radar (GPR) has been widely used as a non-destructive technique to detect subsurface objects. Manual interpretation of GPR data is tedious and time-consuming. To address this challenge, an automatic method based on deep instance segmentation framework developed segment object signatures from scans. The proposed develops the Mask Scoring R-CNN (MS R-CNN) architecture by introducing novel anchoring scheme. By analyzing characteristics hyperbolic objects in scans, set anchor...

Ground penetrating radar (GPR), as a non-invasive instrument, has been widely used in the civil field. The interpretation of GPR data plays vital role underground infrastructures to transfer raw interested information, such diameter. However, diameter identification objects B-scans is tedious and labor-intensive task, which limits further application field environment. paper proposes deep learning-based scheme solve issue. First, an adaptive target region detection (ATRD) algorithm proposed...

Ground Penetrating Radar (GPR) is a non-destructive detection technology based on electromagnetic waves. However, the presence of clutter during data acquisition process has serious impact quality GPR B-scan images. In particular, when strong such as rebar mesh present in surface layer road area being scanned, it can be challenging to locate and identify defects voids faults acquired To suppress recover disturbed defect echoes, we designed an unsupervised SuppRebar-GAN network. Meanwhile,...

Ground-penetrating radar (GPR), as a non-invasive instrument, has been widely used in civil engineering. In GPR B-scan images, there may exist random noise due to the influence of environment and equipment hardware, which complicates interpretability useful information. Many methods have proposed eliminate or suppress noise. However, existing an unsatisfactory denoising effect when image is severely contaminated by This paper proposes multi-scale convolutional autoencoder (MCAE) denoise...

Gesture recognition that is based on high-resolution radar has progressively developed in human-computer interaction field. In a recognition-based system, it challenging to recognize various gesture types because of the lacking transversal feature. this paper, we propose an integrated system frequency modulated continuous wave MIMO combined with deep learning network for recognition. First, pre-processing algorithm, which consists windowed fast Fourier transform and intermediate-frequency...

Antenna array synthesis is an important issue in MIMO radars. By judiciously designing antenna positions, one can create a very long virtual steering vector with small number of antennas and therefore achieve high spatial resolution at cost. This paper presents combinatorial methodology based on cyclic difference sets (CDSs) for minimum redundancy (MR) which seeks to maximize the aperture given antennas. First, key features CDSs CDS-based MR-MIMO layouts are described. Then, analytical...

In this paper, we propose a new ground penetrating radar (GPR) imaging technique based on multi-scale weighted back projection (BP) processing. Firstly, the whole region is discretized by large scale and low-resolution result obtained using traditional BP technique. Secondly, potential targets regions (PTR) are delineated from intensity detection method. PTR, small discretization implemented higher resolution A weight factor designed analyzing statistical characteristics of scattering data...

An improved back projection imaging algorithm for subsurface target detection is presented in this paper. Firstly, the characteristic of scattering data at each time-delay curve traditional procedure analyzed. Secondly, a weight factor designed focal point and an presented. Thirdly, simulation processed. The results both real ground penetrating radar show effectiveness algorithm.

Saliency map generation in synthetic aperture radar (SAR) imagery has become a promising research area, since it close relationship with quick potential target identification, rescue services, etc. Due to the multiplicative speckle noise and complex backscattering SAR imagery, producing satisfying results is still challenging. This paper proposes new saliency approach for using Bayes theory heterogeneous clutter model, i.e., G 0 model. With theory, ratio of probability density functions...

This paper deals with the antenna mutual coupling effect in interferometric aperture synthesis radiometers (IASRs), which degrades system radiometric performance. First, conventional impedance (CMI) is adopted to analyze on performance of IASR and a practical model coupled visibilities developed. Based model, an external calibration method then proposed compensate for effect. In this method, measured are decoupled through difference measurement between original scene naturally occurring...

This paper addresses the issue of how to properly arrange transmitting and receiving (Tx/Rx) antenna elements MIMO radar in order achieve best spatial resolution. By combining positive features cyclic difference sets (CDSs) genetic algorithm (GA), a hybrid optimization method is proposed for optimal design arrays. Numerical results are also presented, showing variety array configurations notable improvement

Convolutional Neural Network- (CNN-) based land cover classification algorithms have recently been applied in hyperspectral images (HSI) field. However, the large-scale training parameters bring huge computation burden to CNN and spatial variability of spectral signatures leads relative low accuracy. In this paper, we propose a CNN-based framework that extracts square matrix representation-based spectral-spatial features performs classification. Numerical results on popular datasets show our...

In this paper, the antenna array design problem in minimum redundancy MIMO radars is addressed, which seeks to maximize virtual aperture for a given number of transmitting and receiving (Tx/Rx) elements. A hybrid optimization method combining cyclic difference sets (CDSs) with simulated annealing (SA) proposed optimal Tx/Rx arrays. Numerical results are also presented, showing large variety configurations notable improvement spatial resolution.

A novel NUSS-FFT(non-uniform spatial sampling FFT) SAR(synthetic aperture radar) imaging algorithm is presented in this paper. It aims at SAR from non-uniform data. Firstly, a brief analysis of echo signal presented. Secondly, the azimuth processing NUSS data analyzed and NUSS-FFT algorithm's procedure Thirdly, simulation processed. Simulation results show effectiveness algorithm.

When using ground-penetrating radar (GPR) detects targets in complex underground scene, due to the nonuniformity of medium, disturbance echo signal at aspects propagation path and attenuation leads phenomenon such as dislocation, missing, distortion characteristic scattering curve GPR record profile. Traditional methods target localization based on hyperbolic vertex detection is no longer valid. In this paper, a method automatic threshold edge template matching was proposed reduce influence...

Radar-based dynamic gesture recognition has attracted great interest of researchers in the field human-computer interaction. A very important problem based on radar is echo data preprocessing technique, it can significantly improve hand accuracy. In this paper, we present a Threshold-Denoising and CA-CFAR (TD-CF) method FMCW for recognition. We first perform 2D-FFT with window function matrix to deal spectrum leakage get Range Doppler Map. An improved wavelet threshold algorithm designed...

Cameras are widely used in many scenes such as robot positioning and unmanned driving, which the camera calibration is a major task this field. The interactive method based on plane board becoming popular due to its stability handleability. However, most methods choose suggestions subjectively from fixed pose dataset, error-prone limited for different models. In addition, these do not provide clear guidelines how place specified pose. This paper proposes new method, named ‘Calibration...

Ground-Penetrating Radar (GPR) is a non-destructive sensing technology that utilizes high electromagnetic frequencies. However, mutual interference waves caused by multiple scattering between targets can significantly complicate the interpretation of GPR B-scan images, especially when shallow obscure deeper ones. Existing methods primarily focus on extracting target signals from background clutter, frequently overlooking impact interference. This paper proposes convolutional neural network,...