Histogram based fuzzy Kohonen clustering network for image segmentation

It is always advisable to apply filtering only on corrupted pixels of images leaving untouched the uncorrupted ones to preserve image features and to avoid blurring effects. We present an algorithm which can detect the corrupted pixels in texture images. It uses a fuzzy Kohonen clustering network that integrates with the fuzzy c-means (FCM) model utilizing the updating strategies of the first and the learning rate of the second.

By integrating Fuzzy Kohonen Clustering Network (FKCN) with Fuzzy Dempster-Shafer Evidential Reasoning Theory (FDSERT), a new multi-source data fusion of Remote Sensing information algorithm is proposed in this paper. The new algorithm can be applied in classification of remote sensing image through FKCN learning and FDSERT fusing. Experimental results comparing with the FKCN algorithm indicates that the classification algorithm of multi-source data fusion of Remote Sensing is superior to that of FKCN algorithm. And the algorithm can obviously improve classification accuracy. At the same time, the algorithm can makes the best of expert knowledge. Therefore the algorithm is an effective classification algorithm of Remote Sensing image.

The clustering of high-dimensional data presents a critical computational problem. Therefore, it is convenient to arrange the cluster centres on a grid with a small dimensional space that reduces computational cost and can be easily visualized. Moreover, in real applications there is no sharp boundary between classes, real datasets are naturally defined in a fuzzy context. Therefore, fuzzy clustering fits better for complex real datasets to determine the best distribution. Self-organizing map (SOM) technique is appropriate for clustering and vector quantization of high-dimensional data. In this paper we present a new fuzzy learning approach called FMIG (fuzzy multilevel interior growing self-organizing maps). The proposed algorithm is a fuzzy version of MIGSOM (multilevel interior growing self-organizing maps). The main contribution of FMIG is to define a fuzzy process of data mapping and to take into account the fuzzy aspect of high-dimensional real datasets. This new algorithm is able to auto-organize the map accordingly to the fuzzy training property of the nodes. In the second step, the introduced scheme for FMIG is clustered by means of fuzzy C-means (FCM) to discover the interior class distribution of the learned data. The validation of FCM partitions is directed by applying six validity indexes. Superiority of the new method is demonstrated by comparing it with crisp MIGSOM, GSOM (growing SOM) and FKCN (fuzzy Kohonen clustering network) techniques. Thus, our new method shows improvement in term of quantization error, topology preservation and clustering ability.

the difficulties of wireless sensor data fusion is to switch the sensor cluster state, namely the problem at some point the direction to which the sensor data fusion. First of all, by the acquisition of knowledge in rough set theory, the typical clustering distribution of multiple sensors can be fused as sample space form decision table data fusion distribution, then according to the method of knowledge reduction in rough sets, attributes and sample data redundancy to delete. And then the clustering analysis and fuzzy Kohonen clustering network, and finally the formation of distribution rules of sensor data fusion. The experimental results show that the model has good classification efficiency, can realizes the clustering distribution of judgment.

Information on the spatial variation of soil fertility attributes is an essential input in precision agriculture for soil management decisions. In this study, soil fertility was evaluated through the spatial distribution of property maps and subsequent integration into fertility classes, as a fundamental basis for the implementation of fertilization plans and amendments adjusted to crop requirements. For the evaluation of fertility, a systematic surface sampling was carried out at 70 sites in the "Agronomy" production field of Romulo Gallegos University, The Castrero sector, Roscio municipality, Guárico state, Venezuela. Ten soil attributes were analyzed: pH, electrical conductivity, organic matter, phosphorus, potassium, calcium, magnesium, and the relative amounts of sand, silt, and clay. Soil property maps were produced by geostatistical analysis and ordinary kriging interpolation, and artificial intelligence techniques based on an artificial neural network classification system, with the FKCN (fuzzy Kohonen clustering network) algorithm, were applied to generate soil fertility classes. The reliability of the maps for each variable was obtained by cross-validation with a reliability of more than 90%. The integration of the maps produced a map composed of five categories. The final soil class model presented a reliability equivalent to 86%, indicating a high degree of homogeneity within the soil classes obtained. This approach overcomes the limitations of traditional methods by integrating multiple variables into a coherent model and is capable of generating information that can be used as a basis for the establishment of experimental plots for research purposes and the specific management of nutrients present in the soil resource of the area under consideration.

Mapping adaptive fuzzy Kohonen clustering network onto distributed image processing system

Mining event logs for knowledge discovery based on adaptive efficient fuzzy Kohonen clustering network

The authors propose a fuzzy Kohonen clustering network which integrates the fuzzy c-means (FCM) model into the learning rate and updating strategies of the Kohonen network. This yields an optimization problem related to FCM, and the numerical results show improved convergence as well as reduced labeling errors. It is proved that the proposed scheme is equivalent to the c-means algorithms. The new method can be viewed as a Kohonen type of FCM, but it is self-organizing, since the size of the update neighborhood and the learning rate in the competitive layer are automatically adjusted during learning. Anderson's IRIS data were used to illustrate this method. The results are compared with the standard Kohonen approach. >

This article presents a fast self-localization method based on ZigBee wireless sensor network and laser sensor, an obstacle avoidance algorithm based on ultrasonic sensors for a mobile robot. The positioning system and positioning theory of ZigBee which can obtain a rough global localization of the mobile robot are introduced. To realize accurate local positioning, a laser sensor is used to extract the features from environment, then the environmental features and global reference map can be matched. From the matched environmental features, the position and orientation of the mobile robot can be obtained. To enable the mobile robot to avoid obstacle in real-time, a heuristic fuzzy neural network is developed by using heuristic fuzzy rules and the Kohonen clustering network. The experiment results show the effectiveness of the proposed method.

Breast cancer has become the primary factor threatening women's health. Automated breast volume scanner (ABVS) is applied for automatic scanning which is meaningful for the rapid and accurate detection of breast tumor. However, accurate segmentation of tumor regions is a huge challenge for clinicians from the ABVS images since it has the large image size and low data quality. Therefore, we propose a novel 3D deep convolutional neural network for automatic breast tumor segmentation from ABVS data. The structure based on 3D U-Net is designed with attention mechanism and transformer layers to optimize the extracted image features. In addition, we integrate the atrous spatial pyramid pooling block and the deep supervision for further performance improvement. The experimental results demonstrate that our model has achieved dice coefficient of 76.36% for 3D segmentation of breast tumor via our self-collected data.

Deep expectation-maximization network for unsupervised image segmentation and clustering

Depth profiling has been performed by using Auger electron spectrometry (AES) and X-ray photoelectron spectrometry (XPS) in combination with Ar-ion sputtering. The data obtained by both surface-analytical methods have been evaluated by means of factor analysis and afterwards by applying an artificial neural network or fuzzy clustering in order to determine the compositional layering of different samples such as a Cr2O3/CrN sandwich layer, tarnish layers on a nickel based alloy and on steel, and the coating of a Si3N4 ceramic powder. The applied artificial neural network was a Kohonen network. It turned out that the method of fuzzy c-means clustering was more successful than Kohonen network due to the fact that fuzzy c-means clustering starts with more input information which can be obtained from factor analysis.

Summary The goal of different methods for clustering of seismic attributes has been to analyze the discrimination ability of the chosen set of attributes. When enough wells are drilled, 3D log properties can be modeled by calculating the acoustic impedance cube. Estimation of log properties from complex seismic attributes becomes very important when only one well is drilled. In this paper, using a subset of complex seismic attributes and the data from one well, sonic log values were estimated in the location of the second well and compared with the actual values of the sonic log at that location (cross-validation). Kohonen’s self organizing maps (SOM or Kohonen’s clustering network) are well known for cluster analysis (unsupervised learning). This class of algorithms is a set of heuristic procedures that suffers from several major problems. A new unsupervised method was used by integrating the fuzzy c-means clustering and Gustafson-Kessel algorithms into the learning and updating strategies of the Kohonen clustering network and attributes are classified using this technique. Embedding the centers of clusters in the Local Linear Modeling Neural Network’s hidden layers leads to a powerful and very fast tool for extrapolation of log properties. Therefore using the minimum possible log data, a method is proposed which can estimate the log values in areas far from the drilled well with a reasonable accuracy. For clustering the attribute data, a few traces are used from different locations in the 3D seismic data; therefore the training time is much less than the conventional methods. The results of modeling, tested on a real data set, confirm the robustness and accuracy of this method compared to conventional tools.

In this article, we present a popular loseless compression/decompression algorithm, GZIP, and the study to implement it on a FPGA based architecture. The algorithm is loseless, and applied to 'bi-level' images of large size. It insures a minimum compression rate for the images we are considering. The proposed architecture for the compressor is based ona hash table and the decompressor is based on a parallel decoder of the Huffman codes.

Segmentation and Classification Models Validation Area Mapping of Peat Lands as Initial Value of Fuzzy Kohonen Clustering Network