A novel RBF neural network for radar emitter recognition based on Rough Sets

With the spreading of radar emitter technology, it is more difficult for traditional methods to recognize radar emitter signals. In this article, a new method is proposed to establish a novel radial basis function (RBF) neural network for radar emitter recognition based on Rough Sets theory. First of all, radar emitter signals describing words are processed by Rough Sets, and the importance weight of each attribute is obtained and the classification rules are extracted. The classification rules are the basis of initial centers of Rough k-means. These initial centers can reduce the computational complexity of Rough k-means efficiently because of a priori knowledge from Rough Sets. In addition, basis functions of neural units of an RBF neural network are improved with attribute importance weights based on Rough Sets theory. The novel network structure makes the RBF neural network more effective. The simulation results show that novel RBF neural network radar emitter recognition can recognize radar emitter signals more effectively than a traditional RBF neural network, because of the improved Rough k-means and the network structure with attribute importance weights.     

多方法融合技术在多目标位置检测中的应用

提出了将多种技术融合一体定位多目标的方法。首先根据计算机视觉原理，采用正交光轴的双摄像机配置检测系统；然后建立去除复杂实际背景和自适应二值化的动态图象预处理模型，并根据矩法搜索和求取目标质心；再利用神经网络模拟人眼感知事物的功能，确立图象点与空间点的映射关系，并将目标质心象点坐标作为输入节点，采用两个BP网并行处理左右两幅图象，快速求得目标空间位置。实践表明，此方法具有一定的实用性，并且在实际应用中取得了满意的效果。     

Multicriteria second‐order neural networks approach to imaging through turbulence

Atmospheric turbulence can greatly limit the spatial resolution in optical images obtained of space objects when imaged with ground‐based telescopes. Two widely used algorithms to remove atmospheric turbulence in this class of images are blind de‐convolution and speckle imaging. Both algorithms are effective in removing atmospheric turbulence, but they use different types of prior knowledge and have different strengths and weaknesses. We have developed a multicriteria cross entropy minimization approach to imaging through atmospheric turbulence and a second‐order neural network implementations. Our simulations illustrated the efficiency of our method. © 2003 Wiley Periodicals, Inc. Int J Imaging Syst Technol 13, 146–151, 2003; Published online in Wiley InterScience (www.interscience.wiley.com). DOI 10.1002/ima.10037     

Radar imaging

Radar systems combining coherent signals with frequency and angular diversity offer the possibility of synthesizing images of complex objects with spatial resolution of a few wavelengths. The availability of high-quality microwave sources and components, high-speed digital computers, and efficient signal-processing algorithms allows radar imaging to be implemented in laboratory environments using commercially-available equipment. The paper summarizes fundamental issues by addressing conceptual and practical limits of radar imaging and presents examples obtained from results of measurements in a laboratory environment. Implementation details of sophisticated operational imaging radars are not covered.©1993 John Wiley & Sons Inc     

Radar-echo tracking by use of invariant moments

A number of techniques to track rainfall patterns by use of radar observations have been developed over the years. We present a method for radar-echo tracking based on Hu invariant moments. The method has been tried on several sequences of test images, and the derived displacement fields were in good agreement with the real motions of the tested objects. For the real data obtained from the conventional meteorological radar in Legionowo the method occasionally failed when changes in the radar echo between observations were too large.     

Gauss Gradient and SURF Features for Landmine Detection from GPR Images

Recently, ground-penetrating radar (GPR) has been extended as a well-known area to investigate the subsurface objects. However, its output has a low resolution, and it needs more processing for more interpretation. This paper presents two algorithms for landmine detection from GPR images. The first algorithm depends on a multi-scale technique. A Gaussian kernel with a particular scale is convolved with the image, and after that, two gradients are estimated; horizontal and vertical gradients. Then, histogram and cumulative histogram are estimated for the overall gradient image. The bin values on the cumulative histogram are used for discrimination between images with and without landmines. Moreover, a neural classifier is used to classify images with cumulative histograms as feature vectors. The second algorithm is based on scale-space analysis with the number of speeded-up robust feature (SURF) points as the key parameter for classification. In addition, this paper presents a framework for size reduction of GPR images based on decimation for efficient storage. The further classification steps can be performed on images after interpolation. The sensitivity of classification accuracy to the interpolation process is studied in detail.     

On the use of time series techniques in shape recognition applications

In this paper, a method for classifying objects based on the use of autoregressive model parameters which are obtained from a time series representation of shape boundaries in digital images of objects is presented. This technique is insensitive to size and is rotation invariant. The objects chosen are four types of aircraft from a digital photograph. Recognition accuracies of more than 90% were obtained for all the pattern classes. All pattern recognition problems involve two random variables, the pattern vector and the class to which it belongs. The interdependence of the two variables is given by the conditional probability density function. The degree of dependence between the pattern vector and the particular class is measured by the “distance”. A simple Bhattacharya distance classifier was used for the purpose.     

基于向量小波与神经网络的图像融合算法

提出了一种基于向量小波和神经网络的图像融合算法.首先对各源图像进行向量小波变换,根据变换后系数计算出各子块图像的清晰度,选取子块图像部分区域清晰度作为前溃神经网络的训练样本,调整神经网络权重;然后用训练好的神经网络组合融合图像的向量小波系数,对组合后的系数进行一致性校验;最后对该系数进行向量小波逆变换,得到融合图像.仿真实验表明,该算法能够较好地解决多传感器图像融合问题,生成的融合图像效果优于有代表性的图像融合方法.     

Advanced image-processing technique for real-time interpretation of ground-penetrating radar images

This work presents the development of an advanced image analysis technique capable of locating buried objects by ground-penetrating radar (GPR) images. The technique requires only a small amount of operator intervention and is fast enough to provide quasi-engineering drawings in real time. The work is intended to simplify the interpretation of the complex pattern found in GPR images. A theoretical development of the method is presented. Results on synthetic and real images from different buried objects are presented, and errors < 7% on the object position are observed with laboratory test objects. © 1998 John Wiley & Sons, Inc. Int J Imaging Syst Technol, 9, 51–59, 1998     

Construction of fuzzy relations by means of fourier holography

Optical Fourier holography is interpreted as a method of creating and recording fuzzy relations between sets represented in the form of images. The properties of the fuzzy relations created by means of a holographic correlator and images implemented in neural network algorithms are considered. Translated from Izmeritel'naya Tekhnika, No. 4, pp. 47–50, April, 1999.     

The identification of objects in active-passive radar systems

Methods of identifying objects in active-passive radar systems from the results of direction finding and a measurement of coordinates are analyzed and the efficiency of the identification is estimated. __________ Translated from Izmeritel’naya Tekhnika, No. 6, pp. 43–48, June, 2008.     

Image fusion for synthetic aperture radar and multispectral images based on sub-band-modulated non-subsampled contourlet transform and pulse coupled neural network methods

Fusion of synthetic aperture radar (SAR) and multispectral (MS) images can contribute to a better visual perception of the objects observed. Unfortunately, many classical approaches have been proven to be unsuitable for this task due to their intrinsic differences in imaging mechanism. In the non-subsampled contourlet transform domain, an alternative fusion method based on pulse coupled neural networks is proposed. To control the amount of SAR features to be integrated into MS image, a gradient-threshold combined modulation is designed for modulating the SAR sub-band coefficients. Experiments demonstrate that the proposed method outperforms its counterparts in spectral preservation and feature enhancement.     

Neural Network Based Thickness Estimation from Multiple Radiographic Images

Back propagation (BP) type artificial neural networks (ANN) have been trained and used for thickness estimations from radiographic images. Test objects have been assembled from different materials and radiographic images of the test objects were obtained for thickness estimations. While some of the study has been based on the synthetic images formed through the radiographic simulation program XRSIM, the rest of the study has used actual radiographic images. The average estimation errors were 7% and 9% when two and three synthetic radiographic images obtained at different x-ray tube settings were used. With the actual images, the thickness of only one of the materials has been estimated and the material was identified. This has been due to the fact that scattering of x-rays by the test object results in a non uniform gray scale variation in the radiographic images even though the object thickness is uniform.     

Performance analysis of neural networks for classification of medical images with wavelets as a feature extractor

In this article, we analyze the performance of artificial neural network, in classification of medical images using wavelets as feature extractor. This work classifies the mammographic image, MRI images, CT images, and ultrasound images as either normal or abnormal. We have tested the proposed approach using 50 mammogram images (13 normal and 37 abnormal), 24 MRI brain images (9 normal and 15 abnormal), 33 CT images (11 normal and 22 abnormal), and 20 ultrasound images (6 normal and 14 abnormal). Four kind of neural network models such as BPN (Back Propagation Network), Hopfield, RBF (Radial Basis Function), and PNN (Probabilistic neural network) were chosen for study. To improve diagnostic accuracy, the feature extracted using wavelets such as Harr, Daubechies (db2, db4, and db8), Biorthogonal and Coiflet wavelets are given as input to the neural network models. Good classification percentage of 96% was achieved using the RBF when Daubechies (db4) wavelet based feature extraction was used. We observed that the classification rate is almost high under the RBF neural network for all the dataset considered. © 2015 Wiley Periodicals, Inc. Int J Imaging Syst Technol, 25, 33–40, 2015     

Spatial resolution enhancement of ultrasound images using neural networks

Spatial resolution in modern ultrasound imaging systems is limited by the high cost of large aperture transducer arrays, which require a large number of transducer elements and electronic channels. A new technique to enhance the spatial resolution of pulse-echo imaging systems is presented. The method attempts to build an image that could be obtained with a transducer array aperture larger than that physically available. We consider two images of the same object obtained with two different apertures, the full aperture and a subaperture, of the same transducer. A suitable artificial neural network (ANN) is trained to reproduce the relationship between the image obtained with the transducer full aperture and the image obtained with a subaperture. The inputs of the neural network are portions of the image obtained with the subaperture (low resolution image), and the target outputs are the corresponding portions of the image produced by the full aperture (high resolution image). After the network is trained, it can produce images with almost the same resolution of the full aperture transducer, but using a reduced number of real transducer elements. All computations are carried out on envelope-detected decimated images; for this reason, the computational cost is low and the method is suitable for real-time applications. The proposed method was applied to experimental data obtained with the ultrasound synthetic aperture focusing technique (SAFT), giving quite promising results. Realtime implementation on a modern, full-digital echographic system is currently being developed.     

Features of the procedure for measuring the amplitude backscattering diagrams of very large objects using ground radar measuring systems

A procedure for measuring the backscattering diagrams of large objects is considered, which reduces the components of the error due to the long-term instability of the parameters of radar measuring stations, when using a calibration without removing the object being investigated from the operating region of the measuring system. The results of an experimental check of the procedure are presented. __________ Translated from Izmeritel’naya Tekhnika, No. 6, pp. 48–50, June, 2008.     

Intelligent sensors using neural networks: the example of amicrosystem for visual inspection

The aim of this article is to show how artificial neural networks and 3D packaging technology have a major role to play in the development of microsystems. A visual inspection system for real-time identification of objects in a scene is described. The system comprises a CMOS or CCD imager, an analogue preprocessing stage that includes a learning mechanism for adapting the system to images of different contrast, and a neural classification stage. The detection of a matrix code using as the classifier a vector support machine is illustrated. As the latter is difficult to realise in VLSI the author has turned to the threshold neural network `Offset', which constructs a parity machine, i.e. a network comprising a single layer of neurons, the output being obtained with the help of a simple exclusive-OR logic gate. Unfortunately the parity machine suffers from overtraining, as the OffSet algorithm converges to a zero error over the entire training base. Nevertheless, if good implementation strategies are available, it is possible to improve the performance in general by combining a large number of classifiers by majority voting. A CMOS VLSI circuit, called SysNeuro, has been fabricated which integrates a parity machine in a square systolic architecture of 4×4 processors. This circuit has variable precision. The number of neurons has been increased by combining 4 SysNeuro chips in a multichip module and stacking three of the modules to form a 3D structure-SysNeuro3D     

融合注意力机制的多尺度深度网络的逆半调方法

目的 运用现有的逆半调方法恢复的图像存在着半色调网纹去除不够理想、图像细节恢复不够清晰等问题,为了进一步提高逆半调图像在平滑区域和纹理细节方面的质量,提出一种基于融合注意力机制的多尺度卷积神经网络的逆半调方法。方法 首先,根据半色调图像网点噪声多频分布特点,设计多尺度卷积网络为基础结构的深度学习网络,从多个不同的尺度抑制半色调网纹并恢复不同尺度的图像信息;然后,应用注意力机制重建图像信息,从而生成逆半调图像;最后,提出多任务损失函数加速网络优化,更好地实现逆半调。结果 实验结果表明,运用此方法得到的逆半调图像在视觉上与原始图像更为相近,恢复出的图像细节更好;在客观评价方面,通过与现有的最先进的方法相比,峰值信噪比平均值提高了0.562～10.095 dB,结构相似度平均值提高了0.01～0.171。结论 该方法可以实现半色调图像的高质量恢复。     

Photon-limited image classification with a feedforward neural network

An application of neural networks to the classification of photon-limited images is reported. A three-level feedforward network architecture is employed in which the input units of the network correspond to the pixels of a two-dimensional image. The network is trained in a minicomputer by the use of the backpropagation technique. The statistics of the network components are analyzed, resulting in a method by which the probability of correct classification of a given input image can be calculated. Photon-limited images of printed characters are obtained with a photon-counting camera and are classified. The experimental results are in excellent agreement with theoretical predictions.     

The possibility of interpreting the results of measurements of the radar characteristics of objects of complex shape using mathematical models

It is shown, using the example of two actual objects, that it is possible to interpret unclear and difficult-to-explain results of measurements of radar characteristics using methods of mathematical modeling. __________ Translated from Izmeritel’naya Tekhnika, No. 3, pp. 44–46, March, 2008.