Search

Image inpainting is a process that tries to fill in missing parts of a degraded image or remove objects from an image, but maintain the realistic content in the image. Various methods can be employed to perform inpainting. Here three benchmark methods of which two are conventional and other GAN based are evaluated to check the effectives of the inpainting process, for different class of images. The visual quality evaluation as well as quantification parameter namely the SSIM score was computed. Based on the test score it was seen that the GAN based method was capable of providing visually good restored images with an average SSIM score of 0.9516, which is an indicator on the quality of the image. The GAN based methods can be further enhanced to provide more realistic images by incorporating functions to extract both local and global contextual content.

Reversible data hiding is a technique that embeds additional information into some distortion-unacceptable cover media, such as military images, in a reversible manner so that the original cover image can be restored after extraction of the hidden information. This work extends a recently proposed reversible data hiding (RDH) scheme of Qu et al. which is based on pixel-based pixel- value-ordering(PPVO) and prediction-error expansion. In Qu et al.’s method, each pixel is predicted using its sorted context pixels. In this work, the pixel neighborhood of each pixel is expanded to optimize the embedding performance in terms of capacity- distortion behavior. This can better exploit image redundancy; achieve superior embedding performance and low distortion. Thus, the proposed method is able to embed adequate data into a cover image with limited distortion. The superiority of this predictor is verified through extensive experimental results. The proposed method outperforms prior works in terms of PSNR. The PSNR of a modified image versus its original one is guaranteed to be above 57.0 dB.

Background: There is little evidence of research outcome data or studies into self-reported back pain in Middle Eastern or Gulf region countries. Within HMC there is anecdotal evidence that suggests that back pain in nurses working in critical care environments accounts for a significant percentage of all sick cause leave. This has the potential to impact on productivity, patient care and quality of life in nurses working within critical care in environments. Methods: A cross center mixed methodology study looking at back pain in nurses working in critical care environments data collection included demographic, occupational, and health characteristics and a Likert questionnaire. This comprised of 10 questions relating to manual handling education, equipment, staffing levels and ergonomics and was distributed in critical care and emergency departments across five hospital sites responses rate n = 450. Results: Outcome data identified 65% of the respondents reported experiencing back pain over the last year with subsequent negative impact on quality of life. There was statically significant association between gender and quality of life p = 0.001; with more women verbalizing a negative impact on quality of life secondary to back pain. There were also statically significant relationships between age with younger age group reporting negative impact on quality of life p = 0.001 and length of time employed in HMC with respondents who have been employed between 1–5 years experiencing the greatest impact p = 0.001. Conclusions: Back pain is an under reported occupational health concern that impacts on productivity and quality of life. Mandatory training in manual handling should feature in corporate educational agenda. Occupational health initiatives' are essential in the management of chronic health conditions.

Reversible data hiding is a technique to embed additional message into some distortion-unacceptable cover media, such as military or medical images, in a reversible manner so that the original cover content can be perfectly restored after extraction of the hidden message. This paper presents a new method using histogram shifting that aims at reducing the embedding distortion and improving watermark payload. The embedding capacity of a histogram-based reversible data hiding technique is primarily determined by the peak height of the histogram. Recently, some studies have tried to embed data in the histogram of prediction errors by modifying the error values. Previous methods insert data only in smooth areas. But by applying the histogram modulation to the image predictionerrors and by considering their immediate pixel neighborhood, the method we propose inserts data in textured areas also. The proposed method, which can be applied to both natural and medical images, outperforms prior works in terms of payload and PSNR.

The concept of data hiding has made it possible to effectively transmit data in such a way that there is no compromise in security. Recent 3D steganographic techniques focus on maximizing the amount of data that can be embedded within the cover model and minimal distortion. A technique based on vertex decimation was recently published, which focuses on minimizing the distortion ensued during data embedding. Since the very aim of a steganographic technique is to hide data, this point is extremely relevant. While this technique is effective, it is subject to high time complexity. This paper proposes an improved technique to finding the referencing neighbors, which is the very aim of vertex decimation, in a less complex manner, when compared to the vertex decimation algorithm proposed in [1].

As E-Learning Systems are gaining more and more popularity; this paper proposes an adaptive E-Learning System based on semantic web technologies. Initially we calculate the users degree of interest towards different course topics based on SMS triggered E-Assessments which could easily be conducted in classrooms or in distant mode. We then classify the different users based on the assessment score. Now it is possible to calculate the semantic similarity between the user evaluations of E-Assessments based on concepts in domain ontology. According to the similarity measures between different users it is possible to classify them in different clusters and finally adaptive recommendations for individual users can be provided based on nearest neighborhoods in the form of resources, assignments and further E-Assessments.