The Influence of Leadership Style on the Effectiveness of Office Management in Integrated Services Business Offices

The aim of this analysis was to ascertain the degree to which the Head of the Integrated Business Services Office's Leadership Style affects the efficacy of the Integrated Business Services Office's management. In data analysis, a descriptive study of one variable was performed to determine the percentage trend of each test variable; second, the commodity moment correlation formula was used to analyze variable relationships; and third, a basic equation model was used to analyze the impact relationship between the independent and dependent variables. The data for this study were processed using a micro statistical program with the assistance of a computer program. The data analysis and interpretation of the study's hypotheses indicate that there is a significant relationship and influence between the variable leadership style of the Integrated Business Services Office's Head and the effectiveness of the Integrated Business Services Office's management, at the 5% significance level or P = 0, 05, respectively.

Advanced Techniques for Digital Image Watermarking

Digital revolution has made easy the production, distribution and access to the multimedia data. With certain business bene ts, arise the challenge of ownership, authentication and privacy of the data. Multimedia data can now be easily edited and reproduced, resulting in losses and secrecy concerns to the owner of the data. There is a dire need to address these issues, otherwise, the challenges and losses will outweigh the bene ts reaped from the digital age. Digital watermarking is the process of embedding an important message in the carrier (Image, Video). In wa- termarking both message and carrier are important. Watermarking can be used to verify the authenticity of the information and establish ownership of the car- rier. There are two major types of watermarking, (i) Robust (ii) Reversible. In this thesis, three new methods of watermarking are presented, to solve the problems of data hiding and content copyright/ownership protection. One method of Robust watermarking is presented, while 2 methods are developed for reversible watermark- ing of images. Reversible watermarking methods are based on new novel predictors developed for the purpose of watermarking. Therst method of reversible watermarking is based on a new D-Mean predictor. The existing image predictors, Median-edge detector (MED) and Gradient adjusted predictor (GAP) were primarily developed for image compression and were used in reversible watermarking as well. The limitation of compression predictors is the in- ability to use multi-side pixels in the prediction process. In the proposed predictor aim is to exploit local correlation of pixels by using east, west, north and south neigh- bors in the prediction process. The predictor operates around an edge-sensitivity threshold to estimate the direction of the edge. In reversible watermarking methods prediction error (PE) histogram is modeled by a Laplacian distribution. This is because of the spatial redundancy in image pixels. Signi cant improvement of the method is observed for standard images. The surge in the histogram peaks at 0 and short tail of PE for D-Mean con rms the superior performance of the proposed D-Mean predictor over MED and GAP methods. Quantitative measures of predic- tor''s performance are Mean Squared Prediction Error (MSPE) and Entropy of PE. Predictors are compared on the basis of MSPE and entropy pf PE . For all the test images D-Mean yields the least MSPE than MED and GAP. Entropy comparison of PE demonstrates the superiority of the D-Mean predictor. Overall, the average per- formance of D-Mean is also better for both MSPE and PE. The e ectiveness of the D-Mean predictor is validated by incorporation in a 2 stage reversible watermarking method. The obtained results are improved than state of the art. The second method of reversible watermarking is based on a hybrid predictor, de- signed over an enlarged 3 3 neighborhood. The embedding process is divided into four phase representation of the image which allows exploitation of larger prediction context thus enhanced prediction accuracy is obtained. To reduce image distortion at lower capacity payloads, sorting of estimated prediction errors is used, sorting is done with reference to variances of prediction context. For improvement at higher capacity payloads, adaptive embedding is used to determine whether to embed sin- gle or two bits in a given prediction error. The approach is based on decomposing the image into four non-overlapping representations. Each of these representations is watermarked in a separate phase making a total of four phases for embedding a watermark. The order of processing of each phase is exible but should be synchro- nized in encoder and decoder. For simplicity, image is scanned for each of the four phases in top-down and left-right fashion. In each scan, the candidate pixel for em- bedding are predicted using their prediction context and the errors are calculated. Histogram shifting and adaptive embedding is used to increase embedding capacity. Experiments were performed to evaluate the performance of the proposed approach in terms of low distortion in the watermarked image. A new method of Robust watermarking is also presented. The watermark embed- ding problem is modeled as a two-stage optimization problem. In therst stage Genetic algorithm (GA) is used for the selection to appropriate wavelet bands and then in the second stage, Particle swarm optimization (PSO) optimized the water- mark strength for each coe cient of the selected wavelet band. Empirical analysis is also performed for suggesting optimum choices for wavelet family and wavelet depth level for watermarking in the wavelet domain, this helps to get most out of the imperceptibility-robustness tradeoof watermarking paradigm. The proposed wa- termarking method embeds the watermark by decomposing the image using discrete wavelet packet transform. In order to achieve desirable imperceptibility choosing a particular proportion of the total number of wavelet bands without compromising on the robustness can be more useful. Once the optimal wavelet bands are found, watermark strength is optimized for the selected bands using PSO. The method is robust against common image processing attacks, i.e. medianltering, noise addi- tion, JPEG compression and frequencyltering. The dissertation also contains a comprehensive survey of reversible watermarking methods and future directions are listed for further investigations.