In this paper, we propose a novel copy-move forgery detection scheme which can accurately localize duplicated regions with a reasonable computational cost. To this end, a new interest point detector is proposed utilizing the advantages of both block-based and traditional keypoint-based methods. The detected keypoints adaptively cover the entire image, even low contrast regions, based on a uniqueness metric. Moreover, a new filtering algorithm is employed which can effectively prune the falsely matched regions. Considering the new interest point detector, an iterative improvement strategy is proposed. The whole procedure is iterated along with adjusting the keypoints density based on the achieved information. The experimental results demonstrate that the proposed scheme outperforms the state-of-the-art methods using two public benchmark databases.

No abstract

No abstract

Based on the self‐assembly of colloidal particles, diacetyl encapsulated colloidosomes from poly (styrene‐co‐butyl acrylate) latex particles were prepared at two different sintering times, and the effects of release media characteristics, including pH (5, 7 and 9), temperature (10, 25 and 40 °C) and shear rate (0, 50 and 100 s−1) on release profile were investigated. Results showed that the sintering process, pH and shear rate of the release media had a significant (p < 0.05) effect on the release of diacetyl from colloidosome, but temperature had no significant effect on the release profile. A diffusion equation was fitted to the experimental data, and diffusion coefficients of 0.025 × 10−15 to 8.15 × 10−15 m2/s were obtained for diacetyl release at various conditions. Copyright © 2014 John Wiley & Sons, Ltd.

In this study we developed an agent‐based model that simulates the release process of encapsulated flavour in different situations. Our model was built in the NetLogo platform, which is a multi‐agent programming language and modelling environment with a good documentation, user‐friendly environment and visualization abilities. The main aim of the work presented is to provide a model that will easily predict the release of encapsulated flavour. In this work we use the model developed to evaluate the flavour‐release properties (such as release time and rate) dependent upon input parameters. Release data of the encapsulated diacetyl (2,3‐butanedione) from alginate–whey‐protein concentrate microcapsules were then used for model validation. Results show that this model can simulate the release‐flavour process well, and there is a good correlation between experimental and model release data. Results also suggested that the flavour‐release process can easily be predicted using this agent‐based model. Copyright © 2015 John Wiley & Sons, Ltd.