Among them the gray level co - occurrence matrix ( glcm ) and gray gradient co - occurrence matrix ( ggcm ) methods , which attributed to the statistic textural analysis scheme were then chosen to extract the textural features of five kind areas on satellite images . in the second part the principle of classification and bp neural network were introduced . combined with textural features , the improved bp neural network successfully performed on the classification of the satellite images
32.
First the sampled image is preprocessed , then five features are extracted from the image preprocessed based on spatial gray level co - occurrence matrix , at last the method of measuring and analyzing of skin texture is proved valid through the result of test of training , classifying and recognizing for skin texture images based on tfbp network
33.
Based on the experiment datum , the rationality of the selection about cutting parameters is analyzed . the third chapter gives the brief expatiation about the concept of texture . the features of workpiece surface texture are extracted with gray co - occurrence matrix and the disadvantage of this method is pointed out
34.
In this paper , we made an investigation into texture feature extraction and classification based on statistic method and its application in multi - spectral image classification . the research works of this paper have been done as follows : firstly , in order to overcome the weakness of gray level co - occurrence matrix ( glcm ) , a new unsupervised texture segment algorithm , based on multi - resolution model , is presented in this thesis
35.
Discovery of association rules is an important class of data mining whose aim is to capture the co - occurrences of itemsets , the most important thing to do is to find the large itemsets effectively , because this is time consuming and will finally decide the efficiency of algorithms . so now the main study is emphasized on how to find the large itemsets with more and more few time
36.
Discovery of association rules is an important class of data mining whose aim is to capture the co - occurrences of itemsets , the most important thing to do is to find the large itemsets effectively , because this is time - consuming and will finally decide the efficiency of algorithms . so now the main study is emphasized on how to find the large itemsets with more and more few times
37.
Firstly , for the errors of text �� character and word , utilizing neighborship of character or word , check character and word errors by character string co - occurrence probability . secondly , for the errors of syntax of text , according to statistic and analysis of a large - scale contemporary chinese corpus , recognize the predicate focus word and the others sentence ingredient , check the syntax errors . thirdly , for the errors of text �� semanteme , establishing semantic dependency relationship tree based on hownet knowledge , presents a method that based on semantic dependency relationship analysis to compute sentence similarity , check the semantic errors
38.
Because this algorithm only include the first order statistical property , but not take the locations of the modulus extrema into account , the second scheme based on the co - occurrence matrix derived form the discrete wavelet frame modulus extrema is proposed , which includes the partial location information extracted from the co - occurrence matrix of the modulus extrema , and so improves the classification performance
39.
Based on data of sar images which have been pretreated , we apply the gray - level co - occurrence matrix method , and particularly study some texture features used for the classification of sar images , including difference variance difference averages difference entropy contrasts energy s variance sum variances inverse difference moment and correlation etc . furthermore we have abstracted features of sar images
40.
This dissertation deals with the content - based image retrieval ( cbir ) theory and technique ; some new features and tools for more concisely and discriminatingly charactering the content of an image are proposed , such as region - based color histogram , grey - primitive co - occurrence matrix , ratio of centripetal moment , ratio of eccentric moment and ratio of inertial moment . a new modified genetic algorithm is also described in this dissertation , which can upgrade the performance of standard genetic algorithm ( sga ) while used in image segmentation
