This paper presents a qualitative, real-time backlighting positioning sensor for the alignment of an optical beam to a minutely deviated diffraction order axis to increase the power coupling efficiency into a multimode fiber in selective launches. Results show that the technique facilitates the alignment of the lenses to the first diffraction order axis and improves the power coupling efficiency into a multimode fiber.

This paper presents the implementation of a thinned fibre Bragg grating as a fuel adulteration sensor for volatile organic compounds. The proposed sensor can detect upto 10% adulteration of benzene, toluene and xylene: hydrocarbons precisely, whereas traditional methods can detect only upto 20% adulteration. The results obtained from the experiments are verified using Finite Difference Time Domain method. It is found that experimental results have very less deviation from simulation results. The proposed sensor provides us with the new possibility that may have commercial application, as well.

With the advent of technology, video has become a prominent entity that is shared over networks. With easy availability of various editing tools, data integrity and ownership issues have caused great concern worldwide. Video watermarking is an evolving field that may be used to address such issues. Till date, most of the algorithms have been developed for uncompressed domain watermarking and implemented on software platforms. They provide flexibility and simplicity, but at the same time, they are not suited for real-time applications. They work offline where videos are captured and then watermark is embedded in the video. In the present work, a hardware-based implementation of video watermarking is proposed that overcomes the limitation of software watermarking methods and can be readily adapted to the H.264 standard. This paper focuses on an invisible and robust video watermarking scheme, which can be easily implemented as an integral part of the standard H.264 encoder. The proposed watermarking algorithm involves Integer DCT-based watermark embedding method, wherein Integer DCT is calculated with a fully parallel approach resulting in better speed. The proposed video watermarking is designed with pipelining and parallel architecture for real-time implementation. Here, scene change detection technique is used to improve the performance. Different planes of the watermark are embedded in different frames of a particular scene in order to achieve robustness against various temporal attacks.