Search

Objective: The aim of this study is to introduce 23 different binary dissimilarity/distance coefficients determined for different values by using derived data sets and to evaluate them by revealing their properties. Material and Methods: In this study, the dissimilarity/distance coefficients put forward for binary data are considered. In the study, data were derived for 35 different values in the range of using the Phyton-random library. In the derivation of the data, firstly, which cell shown with will be assigned value, then the value to be assigned to the relevant cell was determined. 286 for, 815 for and 1000 different data sets for were used in the study. Results: Although the value range of all dissimilarity/distance coefficients for binary data is expected to be 0 to 1, this range is not valid for all coefficients. Out of 23 different coefficients, 12 take values in this range. According to the hierarchical clustering analysis, most of the dissimilarity/distance coefficients are similar. Conclusion: In general, the values of almost all coefficients increase from a fixed minimum to a fixed maximum as the samples become more dissimilar. However, the Sokal-Michener, Hamming and variance coefficients increase linearly with dissimilarity/distance smoothly for all values. The coefficient suggested by Sokal-Michener stands out among all the coefficients because the value range is 0-1 and is parallel to the dissimilarity/distance increase. Cosine, Hamming, Euclid I and Euclid II coefficients are affected by the number of and other coefficients are not. Therefore, it has been determined that a significant part of the dissimilarity/distance coefficients are independent of the sample size. [ABSTRACT FROM AUTHOR]

Nowadays, a lot of data hiding algorithms for binary images have been proposed in the literature. The main purpose of these algorithms is to provide data hiding with high capacity and high visual quality. The pattern based data hiding algorithms are commonly used to achieve this purpose. In this paper, a new data hiding algorithm is suggested. This algorithm is based on minesweeper game for binary images. Neighborhood degree is calculated by selecting mine value as 0 or 1. The size of 3x3 non overlapping blocks have been used to determine neighborhood degree. Data hiding process is implemented by modifying the obtained neighborhood degree. Hamming distance is calculated to keep distortion in minimum level. In the selection of similar blocks with different neighborhood degrees, privacy of the algorithm is provided by using random number generators. The experimental results showed that the proposed data hiding algorithm resulted successfully. [ABSTRACT FROM AUTHOR]