Classification of multispectral satellite image data using improved NRBF neural networks

Spectral clustering is one of the most popular modern clustering techniques for conventional data. However, the application of the general spectral clustering method in the geostatistical data framework poses a double challenge. Firstly, applied to geostatistical data, the general spectral clustering method produces clusters that are spatially non-contiguous which is undesirable for many geoscience applications. Secondly, it is limited in its applicability to large-scale problems due to its high computational complexity. This paper presents a spectral clustering method dedicated to large-scale geostatistical datasets in which spatial dependence plays an important role. It extends a previous work to large-scale geostatistical datasets by computing the similarity matrix only at a reduced set of locations over the study domain referred to as anchor locations. It has the advantage of using all data during the computation of the similarity matrix at anchor locations; so there is no sacrifice of data. The spectral clustering algorithm can then be efficiently performed on this similarity matrix at anchor locations rather than all data locations. Given the resulting cluster labels of anchor locations, a weighted k-nearest-neighbour classifier is trained using their geographical coordinates as covariates and their cluster labels as the response. The assignment of clustering membership to the entire data locations is obtained by applying the trained classifier. The effectiveness of the proposed method to discover spatially contiguous and meaningful clusters in large-scale geostatistical datasets is illustrated using the US National Geochemical Survey database.

To accomplish the effective classifier and secure the accurate classification capabilities of black plastics, a comprehensive design methodology of fuzzy radial basis function neural networks is developed with the aid of principal component analysis and particle swarm optimization. Plastics recycling is the competitive method which can deal with the shortage of natural resource. To recycle and reuse the waste plastics, this study is given as the key issue to identify and classify waste plastics by resin type such as polyethylene terephthalate, polypropylene, polystyrene, etc. To complement the weak points of recognition and classification of the near-infrared radiation equipment, Raman spectroscopy is used to obtain qualitative as well as quantitative analysis of black plastics. To improve the identification performance of black plastics, an intelligent computing algorithm such as fuzzy radial basis function neural networks classifier and preprocessing algorithm as principal component analysis are applied to analyze and classify the obtained spectrum of black plastics. Finally, to optimize the structure as well as parameters of fuzzy radial basis function neural networks, particle swarm optimization technique is used. The obtained experimental results show that the proposed network architecture exhibits high classification capabilities in practical applications.

This paper uses a radial basis function (RBF) neural network for star identification. A feature vector consisting of distance and angle cosine values is constructed as a library of recognition feature stars and input to the neural network. Then, the network’s output is the number of the corresponding primary star. Comparing the training efficiency and robustness of the grid algorithm with the BP neural network and radial basis function neural network for star identification, the experimental results show that the radial basis function neural network has a strong anti-interference ability against noise and pseudo-star interference. The following conclusions are drawn from the simulation tests: the sample recognition rate of radial basis function neural network algorithm can reach 96% for star position and star addition noise; when the proportion of pseudo-star reaches 20%, the sample recognition rate of radial basis function neural network algorithm is still higher than grid algorithm and BP neural network, which indicates that the radial basis function neural network algorithm has the solid anti-interference ability to both noise and pseudo-star. Therefore, the radial basis function neural network algorithm based on star Identification has specific application prospects.

A survey of kernel and spectral methods for clustering

A neural network-based methodology for time and memory efficient supervised or unsupervised classification in heavily demanding applications is presented in this chapter. Significantly increased speed in the design (training) of neural, fuzzy and statistical classifiers as well as in the classification phase is achieved by: (a) using a self-organizing feature map (SOFM) for vector quantization and indexed representation of the input data space; (b) appropriate training set reduction using the SOFM prototypes followed by necessary modifications of the training algorithms (supervised techniques); (c) clustering of neurons on maps instead of clustering the original data (unsupervised techniques); and (d) fast indexed classification. Finally, a demonstration of this methodology involving the design of multiple classifiers is performed on Land-Cover classification of multispectral satellite image data showing increased speed with respect to both training and classification times.

Digital image recognition of plant diseases could reduce the dependence of agricultural production on the professional and technical personnel in plant protection field and is conducive to the development of plant protection informatization. In order to find out a method to realize image recognition of plant diseases, four kinds of neural networks including backpropagation (BP) networks, radial basis function (RBF) neural networks, generalized regression networks (GRNNs) and probabilistic neural networks (PNNs) were used to distinguish wheat stripe rust from wheat leaf rust and to distinguish grape downy mildew from grape powdery mildew based on color features, shape features and texture features extracted from the disease images. The results showed that identification and diagnosis of the plant diseases could be effectively achieved using BP networks, RBF neural networks, GRNNs and PNNs based on image processing. For the two kinds of wheat diseases, the best prediction accuracy was 100% with the fitting accuracy equal to 100% while BP networks, GRNNs or PNNs were used, and the best prediction accuracy was 97.50% with the fitting accuracy equal to 100% while RBF neural networks were used. For the two kinds of grape diseases, the best prediction accuracy was 100% with the fitting accuracy equal to 100% while BP networks, GRNNs or PNNs were used, and the best prediction accuracy was 94.29% with the fitting accuracy equal to 100% while RBF neural networks were used.

Accurate monitoring of soil salinization plays a key role in the ecological security and sustainable agricultural development of arid regions. As a branch of artificial intelligence, machine learning acquires new knowledge through self-learning and continuously improves its own performance. The purpose of this study is to combine Sentinel-2 Multispectral Imager (MSI) data and MSI-derived covariates with measured soil salinity data and to apply three machine learning algorithms in modeling to estimate and map the soil salinity in the study sample area. According to the convenient transportation conditions, the study area and sampling quadrat were set up, and the 5-point method was used to collect the soil mixed samples, and 160 soil mixed samples were collected. Kennard–Stone (K–S) algorithm was used for sample classification, 70% for modeling and 30% for verification. The machine learning algorithm uses Support Vector Machines (SVM), Artificial Neural Network (ANN), and Random Forest (RF). The results showed that (1) the average reflectance of each band of the MSI data ranged from 0.21–0.28. According to the spectral characteristics corresponding to different soil electrical conductivity (EC) levels (1.07–79.6 dS m−1), the spectral reflectance of salinized soil in the MSI data ranged from 0.09–0.35. (2) The correlation coefficient between the MSI data and MSI-derived covariates and soil EC was moderate, and the correlation between certain MSI data sets and soil EC was not significant. (3) The SVM soil EC estimation model established with the MSI data set attained a higher performance and accuracy (R2 = 0.88, root mean square error (RMSE) = 4.89 dS m−1, and ratio of the performance to the interquartile range (RPIQ) = 1.96, standard error of the laboratory measurements to the standard error of the predictions (SEL/SEP) = 1.11) than those attained with the soil EC estimation models established with the RF and ANN models. (4) We applied the SVM soil EC estimation model to map the soil salinity in the study area, which showed that the farmland with higher altitudes discharged a large amount of salt to the surroundings due to long-term irrigation, and the secondary salinization of the farmland also caused a large amount of salt accumulation. This research provides a scientific basis for the simulation of soil salinization scenarios in arid areas in the future.

Fault diagnosis for wireless sensor is very important to ensure signal acquisition precision. Radial basis function (RBF) neural network has strong classification ability. However, the selection of the connection weights, the hidden centers and the widths has an important influence on the classification performance of the RBF neural network in the learning process of RBF neural network. Thus, ant colony optimization is employ to gain the parameters of radial basis function neural network. Therefore, a novel method for fault diagnosis of wireless sensor based on RBF neural network and ant colony optimization is presented. The results of computational experiments show that ACO- RBF neural network has a great higher than RBF neural network.

The traditional methods to derive sea ice concentration are mainly from low resolution microwave data, which is disadvantageous to meet the grid size requirement of high resolution climate models. In this paper, moderate resolution imaging spectroradiometer (MODIS)/Terra calibrated radiances Level-1B (MOD02HKM) data with 500 m resolution in the vicinity of the Abbot Ice Shelf, Antarctica, is unmixed, respectively, by two neural networks to extract the sea ice concentration. After two different neural network models and MODIS potential open water algorithm (MPA) are introduced, a MOD02HKM image is unmixed using these neural networks and sea ice concentration maps are derived. At the same time, sea ice concentration for the same area is extracted by MPA from MODIS/Terra sea ice extent (MOD29) data with 1 km resolution. Comparisons among sea ice concentration results of the three algorithms showed that a spectral unmixing method is suitable for the extraction of sea ice concentration with high resolution and the accuracy of radial basis function neural network is better than that of backpropagation.

Control chart patterns (CCPs) are the important statistical process control tools for determining whether a process is run in its intended mode or in the presence of unnatural patterns. In this paper we proposed a hybrid intelligent technique for recognition of the CCPs. In this technique we have used a proper set of the shape features and statistical features as the efficient characteristics of the patterns. Then we proposed a hybrid heuristic recognition system based on particle swarm optimization (PSO) to improve the generalization performance of the radial basis function neural network (RBFNN) classifier. For this purpose, we have optimized the classifier design by searching for the best value of the parameters that tune its discriminate function. The obtained results show that the proposed technique has high recognition accuracy in comparison with other works. This recognition accuracy achieves with lower training samples.

Classical radial basis function (RBF) neural network directly projects input samples into a high dimension feature space through some radial basis functions, and does not take account of the high-order statistical relationship among variables of input samples. But the high-order statistical relationship does play an important part in pattern recognition (classification) area. In order to take advantage of the high-order statistical relationship among variables of input samples in neural network, a novel independent radial basis function (IRBF) neural network is proposed in this paper. Then a new hybrid system combining multiresolution analysis, principal component analysis (PCA) and our proposed IRBF neural network is also proposed for face recognition. According to experiments on FERET face database, our proposed approach could outperform newly proposed ICA algorithm. And it is also confirmed that our proposed approach is more robust to facial expression, illumination and aging than ICA in face recognition.

China has been facing serious land degradation and desertification in its north and northwest arid and semi-arid areas. Monitoring the dynamics of percentage vegetation cover (PVC) using remote sensing imagery in these areas has become critical. However, because these areas are large, remote, and sparsely populated, and also because of the existence of mixed pixels, there have been no accurate and cost-effective methods available for this purpose. Spectral unmixing methods are a good alternative as they do not need field data and are low cost. However, traditional linear spectral unmixing (LSU) methods lack the ability to capture the characteristics of spectral reflectance and scattering from endmembers and their interactions within mixed pixels. Moreover, existing nonlinear spectral unmixing methods, such as random forest (RF) and radial basis function neural network (RBFNN), are often costly because they require field measurements of PVC from a large number of training samples. In this study, a cost-effective approach to mapping PVC in arid and semi-arid areas was proposed. A method for selection and purification of endmembers mainly based on Landsat imagery was first presented. A probability-based spectral unmixing analysis (PBSUA) and a probability-based optimized k nearest-neighbors (PBOkNN) approach were then developed to improve the mapping of PVC in Duolun County in Inner Mongolia, China, using Landsat 8 images and field data from 920 sample plots. The proposed PBSUA and PBOkNN methods were further validated in terms of accuracy and cost-effectiveness by comparison with two LSU methods, with and without purification of endmembers, and two nonlinear approaches, RF and RBFNN. The cost-effectiveness was defined as the reciprocal of cost timing relative root mean square error (RRMSE). The results showed that (1) Probability-based spectral unmixing analysis (PBSUA) was most cost-effective and increased the cost-effectiveness by 29.3% 29.3%, 33.5%, 50.8%, and 53.0% compared with two LSU methods, PBOkNN, RF, and RBFNN, respectively; (2) PBSUA, RF, and RBFNN gave RRMSE values of 22.9%, 21.8%, and 22.8%, respectively, which were not significantly different from each other at the significance level of 0.05. Compatibly, PBOkNN and LSU methods with and without purification of endmembers resulted in significantly greater RRMSE values of 27.5%, 32.4%, and 43.3%, respectively; (3) the average estimates of the sample plots and predicted maps from PBSUA, PBOkNN, RF, and RBFNN fell in the confidence interval of the test plot data, but those from two LSU methods did not, although the LSU with purification of endmembers improved the PVC estimation accuracy by 25.2% compared with the LSU without purification of endmembers. Thus, this study indicated that the proposed PBSUA had great potential for cost-effectively mapping PVC in arid and semi-arid areas.

The performance enhancement of system identification of various plastic materials to effectively recycle the waste plastics arises as a key issue studied here. For black plastics, which contain carbon black, one is unable to discriminate it from other materials. To facilitate the identification process, Fourier transform-infrared with attenuated total reflectance is used to carry out qualitative as well as quantitative analysis of black plastics. Since a spectrum obtained in this manner constitutes highly dimensional data, feature reduction becomes necessary to extract sound features and reduce the dimensionality of the original spectrum. In this study, three types of feature extraction techniques are considered: peak detection technique, feature extraction based on the chemical characteristics, and fuzzy transform-based feature extraction to determine sound discriminative features. In order to enhance classification process, fuzzy radial basis function neural networks classifier is constructed; these architectures of the classifiers take advantage of the hybrid technologies. Based upon experimental studies, it is shown that the proposed classification system with the feature extraction techniques exhibits superior performance over the performance reported for the already studied classifiers.

The study compared the forecasting performance of two multiple-input and multiple-output (MIMO) and two single-input and single-output (SISO) neural networks using 4G network traffic data aggregated into daily, weekly and monthly time spans. To explore the best configuration of SISO and MIMO neural networks, the empirical traffic data of 1-input, 2- input and 3-input were used together with varying the parameters of the models. The study concluded that for 2-input, MIMO Radial basis function neural (RBFN) network performed better than the 2-input MIMO Multilayer perceptron (MLP) neural network in predicting the traffic data. In the case of 3-input, MLP network was found to be more efficient than RBFN network. In the scenario of SISO architecture, the MLP network outperformed the RBFN network for 4G daily, weekly and monthly traffic data.

The Sixth International Symposium on Neural Networks (ISNN 2009)