- 标题
- 摘要
- 关键词
- 实验方案
- 产品
-
V-RBNN Based Small Drone Detection in Augmented Datasets for 3D LADAR System
摘要： A common countermeasure to detect threatening drones is the electro-optical infrared (EO/IR) system. However, its performance is drastically reduced in conditions of complex background, saturation and light re?ection. 3D laser sensor LiDAR is used to overcome the problems of 2D sensors like EO/IR, but it is not enough to detect small drones at a very long distance because of low laser energy and resolution. To solve this problem, A 3D LADAR sensor is under development. In this work, we study the detection methodology adequate to the LADAR sensor which can detect small drones at up to 2 km. First, a data augmentation method is proposed to generate a virtual target considering the laser beam and scanning characteristics, and to augment it with the actual LADAR sensor data for various kinds of tests before full hardware system developed. Second, a detection algorithm is proposed to detect drones using voxel-based background subtraction and variable radially bounded nearest neighbor (V-RBNN) method. The results show that 0.2 m L2 distance and 60% expected average overlap (EAO) indexes are satis?ed for the required speci?cation to detect 0.3 m size of small drones.
关键词： 3D sensor,3D LADAR,clustering,drone detection,LiDAR,fusion data
更新于2025-09-10 09:29:36
-
[IEEE 2018 International Conference on Machine Learning and Cybernetics (ICMLC) - Chengdu, China (2018.7.15-2018.7.18)] 2018 International Conference on Machine Learning and Cybernetics (ICMLC) - Spectral-Spatial Sparse Subspace Clustering Based On Three-Dimensional Edge-Preserving Filtering For Hyperspectral Image
摘要： Due to the 3-D property of raw HSI cubes, 3-D spectral-spatial ?lter becomes an effective way for extracting spectral and spatial signatures from HSI. In this paper, a new spectral-spatial sparse subspace clustering framework based on 3-D edge-preserving ?ltering is proposed to improve the clustering accuracy of HSI. First, the initial sparse coef?cient matrix is obtained in the s-parse representation process of the classical SSC model. Then, a 3-D edge-preserving ?ltering is conducted on the initial sparse coef?cient matrix to get a more accurate one, which is used to build the similarity graph. Finally, the clustering result of H-SI data is achieved by employing the spectral clustering algorithm to the similarity graph. Speci?cally, the ?ltered matrix can not only capture the spectral-spatial features but the inten-sity differences. Experimental results demonstrate the poten-tial of including the proposed 3-D edge-preserving ?ltering in-to the SSC framework can improve the clustering accuracy.
关键词： Hyperspectral images (HSIs),Sparse subspace clustering (SSC),3-D edge-preserving ?lters (3-D EPFs),Intensity differences
更新于2025-09-10 09:29:36
-
[IEEE 2018 International Conference on Advances in Computing, Communications and Informatics (ICACCI) - Bangalore, India (2018.9.19-2018.9.22)] 2018 International Conference on Advances in Computing, Communications and Informatics (ICACCI) - Quantum Spider Monkey Optimization (QSMO) Algorithm for Automatic Gray-Scale Image Clustering
摘要： In automatic image clustering, high homogeneity of each cluster is always desired. The increase in number of thresholds in gray scale image segmentation/clustering poses various challenges. Recent times have witnessed the growing popularity of swarm intelligence based algorithms in the field of image segmentation. The Spider Monkey Optimization (SMO) algorithm is a notable example, which is motivated by the intelligent behavior of the spider monkeys. The SMO is broadly categorized as a fission-fusion social structure based intelligent algorithm. The original version of the algorithm as well as its variants have been successfully used in several optimization problems. The current work proposes a quantum version of SMO algorithm which takes recourse to quantum encoding of its population along with quantum variants of the intrinsic operations. The basic concepts and principles of quantum mechanics allows QMSO to explore the power of computing. In QMSO, qubits designated chromosomes operate to drive the solution toward better convergence incorporating rotation gate in Hilbert hyperspace. A fitness function associated with maximum distance between cluster centers have been introduced. An application of the proposed QSMO algorithm is demonstrated on the determination of automatic clusters from real life images. A comparative study with the performance of the classical SMO shows the efficacy of the proposed QSMO algorithm.
关键词： automatic clustering,quantum computing,quantum spider monkey optimization,Clustering,spider monkey optimization
更新于2025-09-09 09:28:46
-
Camera recognition with deep learning
摘要： In this paper, camera recognition with the use of deep learning technique is introduced. To identify the various cameras, their characteristic photo-response non-uniformity (PRNU) noise pattern was extracted. In forensic science, it is important, especially for child pornography cases, to link a photo or a set of photos to a specific camera. Deep learning is a sub-field of machine learning which trains the computer as a human brain to recognize similarities and differences by scanning it, in order to identify an object. The innovation of this research is the use of PRNU noise patterns and a deep learning technique in order to achieve camera identification. In this paper, AlexNet was modified producing an improved training procedure with high maximum accuracy of 80%–90%. DIGITS showed to have identified correctly six cameras out of 10 with a success rate higher than 75% in the database. However, many of the cameras were falsely identified indicating a fault occurring during the procedure. A possible explanation for this is that the PRNU signal is based on the quality of the sensor and the artefacts introduced during the production process of the camera. Some manufacturers may use the same or similar imaging sensors, which could result in similar PRNU noise patterns. In an attempt to form a database which contained different cameras of the same model as different categories, the accuracy rate was low. This provided further proof of the limitations of this technique, since PRNU is stochastic in nature and should be able to distinguish between different cameras from the same brand. Therefore, this study showed that current convolutional neural networks (CNNs) cannot achieve individualization with PRNU patterns. Nevertheless, the paper provided material for further research.
关键词： Forensic sciences,camera identification,individualization,clustering,deep learning
更新于2025-09-09 09:28:46
-
Dynamics of gold nanoparticle clusters observed with liquid-phase electron microscopy
摘要： The dynamics of processes of nanoparticles such as diffusion, attraction and repulsion, and self-assembly of structures of nanoparticles at the solid-liquid interfaces differ significantly from those occurring for bulk conditions and their fundamental physical rules are still unknown. Here, we used liquid phase scanning transmission electron microscopy (LP-STEM) to study several aspects of nanoparticle dynamics of colloidal chitosan coated gold nanoparticle (TCHIT-AuNP) clusters in a liquid layer enclosed between two SiN membranes. We found that upon beam irradiation using an electron flux of 0.9 e?/s?2, the AuNPs assembled in clusters that shifted and rotated with time. The newly formed clusters could join and form larger clusters via a mechanism of oriented attachment. By increasing the electron flux to 6.2 e?/s?2, we observed the fragmentation of some of the clusters and TCHIT-AuNPs were exchanged between clusters. At the highest electron flux studied 25 e?/s?2, we observed AuNPs moving at a very slow speed compared to Brownian motion in liquid even though they were not permanently attached or pinned to the liquid-enclosing membrane. Experiments using branched polyethylenimine (BPEI) coated AuNPs were carried out for comparison.
关键词： nanoparticle clustering,nanoparticle dynamics,solid-liquid interface,gold nanoparticles,scanning transmission electron microscopy,liquid-phase electron microscopy
更新于2025-09-09 09:28:46
-
[IEEE IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - Valencia, Spain (2018.7.22-2018.7.27)] IGARSS 2018 - 2018 IEEE International Geoscience and Remote Sensing Symposium - Fast Airplane Detection with Hierarchical Structure in Large Scene Remote Sensing Images at High Spatial Resolution
摘要： In order to detect airplane efficiently in large scene remote sensing images with high spatial resolution, a hierarchical detection framework is proposed according to the process of locating the target area in the downsampled image and then detecting the target in the original high-resolution image of that target region. First, we locate the airport in the downsampled original high-resolution image based on line features (midpoint coordinate and angle) clustering. A visual saliency model based on Itti model is then applied to detect airplanes candidates in the resulting airport region with the original high-resolution image. Finally, scale invariant feature transformation and support vector machine are used to classify the airplanes candidates. Experimental results indicate that the proposed method presents higher recall rate and detection speed. Meanwhile, the time cost of airport detection is much less than representative methods.
关键词： Large scene image,Straight line clustering,Airplane detection,Airport detection,Saliency map
更新于2025-09-09 09:28:46
-
Dual-Constraint Spatiotemporal Clustering Approach for Exploring Marine Anomaly Patterns Using Remote Sensing Products
摘要： Spatiotemporal clustering patterns of marine anomaly variations are the focus of much current global climate change research. Marine anomaly variations have multidimensional attributes and are spatiotemporally continuous; existing methods for clustering face challenges in mining effectively their spatiotemporal clustering patterns. Using long-term marine remote sensing products, we present the dual-constraint spatiotemporal clustering approach (DcSTCA) for exploring marine clustering patterns. The DcSTCA includes three steps. The first step constructs a spatiotemporal grid cube based on the spatial connectivity and time evolution process of marine anomaly variations, which is used to search the spatiotemporal neighborhood. The second step calculates the proximities in space, time, and thematic attribute to obtain the spatiotemporal clustering cores and spatiotemporal density of each grid cell. The final step derives spatiotemporal clustering patterns by connecting the clustering cores and their spatiotemporal neighbors according to their density connectivity. Experiments on simulated datasets are used to demonstrate the effectiveness and the advantages of the DcSTCA compared with spatial–temporal density-based spatial clustering of applications with noise (ST-DBSCAN). The applications on sea surface temperature in the Pacific Ocean show that the DcSTCA can effectively explore marine clustering patterns from remote sensing products, and these mined clustering patterns may provide new references for global change research.
关键词： the Pacific Ocean,Data mining,marine anomaly variations (MAVs),remote sensing products,spatiotemporal clustering
更新于2025-09-09 09:28:46
-
[IEEE 2018 International Conference on Computation of Power, Energy, Information and Communication(ICCPEIC) - Chennai (2018.3.28-2018.3.29)] 2018 Internat2018 International Conference on Computation of Power, Energy, Information and Communication (ICCPEIC)ional conference on computation of power, energy, Information and Communication (ICCPEIC) - Shrub Ailment Recognization Using Advanced Image Processing
摘要： To increase the productivity of the crop disease detection has to be done. The disease identification is done by normal eye observation method along with the technique of image processing. The objective is to deep to investigate about existing work specific to disease detection in plants. The challenge faced in the naked eye observation method is time consuming and no accuracy, even some types of disease not visible. Challenges faced in image processing is internal structure variation in plant leaf may also due to the reason for lack of humidity, pressure.
关键词： Diseasedetection,K-means clustering algorithm,Image processing,Digital images
更新于2025-09-09 09:28:46
-
Molten Steel Level Detection From Thermal Image Sequence Based on the Characteristics of Adhesive Flux
摘要： High-temperature medium and time-varying covering flux lead to the difficulty of molten steel level measurement. For the measurement, in our previous work, a novel principle by using temperature gradient was proposed by us, and a refractory sensor was inserted into the metallurgical container to sense the temperature gradients of the flux and the molten steel. However, sometimes liquid adhesive flux on the sensor surface disables the extraction of true temperature gradients. To fix this problem, two new models, the adhesion thickness model and the adhesion flowability model, which are inspired by the adhesion mechanism of the flux, are proposed to detect the steel-flux interface. A unified approach, sequential clustering of the shapes of the pixel gray-time curves, is introduced to conduct the detection. On this basis, the thermal image sequence with 4-D spacetime information of the sensor is used for clustering. First, gray values of each pixel in the sequence are sorted in the time dimension, and grouping of the pixels in space dimension is done. Then, the region of interest is extracted from the image sequence to remove the invalid pixels, and sequential clustering is conducted with each group. Finally, the confidence of the clustering results is measured and the clustering results with the confidence higher than the threshold are retained to detect the steel-flux interface. By utilizing the two new models, the standard deviation of the measurement errors reduces from 4.8 to 3.7 mm.
关键词： thermal image sequence,steel-flux interface detection,Adhesion characteristic,4-D spacetime information,sequential clustering
更新于2025-09-09 09:28:46
-
Hyperspectral Tissue Image Segmentation using Semi-Supervised NMF and Hierarchical Clustering
摘要： Hyperspectral imaging (HSI) of tissue samples in the mid-infrared (mid-IR) range provides spectro-chemical and tissue structure information at sub-cellular spatial resolution. Disease-states can be directly assessed by analyzing the mid-IR spectra of different cell-types (e.g. epithelial cells) and sub-cellular components (e.g. nuclei), provided we can accurately classify the pixels belonging to these components. The challenge is to extract information from hundreds of noisy mid-IR bands at each pixel, where each band is not very informative in itself, making annotations of unstained tissue HSI images particularly tricky. Because the tissue structure is not necessarily identical between the two sections, only a few regions in unstained HSI image can be annotated with high confidence, even when serial (or adjacent) H&E stained section is used as a visual guide. In order to completely use both labeled and unlabeled pixels in training images, we have developed an HSI pixel classification method that uses semi-supervised learning for both spectral dimension reduction and hierarchical pixel clustering. Compared to supervised classifiers, the proposed method was able to account for the vast differences in spectra of sub-cellular components of the same cell-type and achieve an F1-score of 71.18% on two-fold cross-validation across 20 tissue images. To generate further interest in this promising modality we have released our source code and also showed that disease classification is straightforward after HSI image segmentation.
关键词： microspectroscopy,semi-supervised learning,hierarchical clustering,Hyperspectral imaging,non-negative matrix factorization
更新于2025-09-09 09:28:46