MUCKE aims to mine a large volume of images, to structure them conceptually and to use this conceptual structuring in order to improve large-scale image retrieval. The last decade witnessed important progress concerning low-level image representations. However, there are a number problems which need to be solved in order to unleash the full potential of image mining in applications. The central problem with low-level representations is the mismatch between them and the human interpretation of image content. This problem can be instantiated, for instance, by the incapability of existing descriptors to capture spatial relationships between the concepts represented or by their incapability to convey an explanation of why two images are similar in a content-based image retrieval framework. We start by assessing existing local descriptors for image classification and by proposing to use co-occurrence matrices to better capture spatial relationships in images. The main focus in MUCKE is on cleaning large scale Web image corpora and on proposing image representations which are closer to the human interpretation of images. Consequently, we introduce methods which tackle these two problems and compare results to state of the art methods. Note: some aspects of this deliverable are withheld at this time as they are pending review. Please contact the authors for a preview.

http://ieeexplore.ieee.org/iel2/4524/12820/00592460.pdf

Image Processing

A motion adaptive deinterlacing algorithm is presented in this paper. It consists of the ELA-median directional interpolation, same-parity 4-field horizontal motion detection, morphological operation for noise reduction and adaptive threshold adjusting. The edges can be sharper when the ELA-median directional interpolation is adopted. The same-parity 4-field horizontal motion detection detects faster motion and makes more accurate determinations about where objects are going to move. The morphological operation for noise reduction and adaptive threshold adjusting preserve the actual texture of the original objects. The proposed method achieves cost-efficient hardware implementation with low complexity, low memory usage, and high-speed processing capability. In addition, it consumes less time in producing high-quality images and allows the audience to enjoy a high-quality TV sequence on their progressive devices. The experimental results show that the proposed algorithm is more cost-effective than previous systems.

/pdf/motion-adaptive-interpolation-with-horizontal-motion-hhamstcs89.pdf

Motion adaptive interpolation with horizontal motion detection for deinterlacing

This paper reviews a range of image and video artefacts, their causes, and mitigation techniques. Due to processing and distortions in transmission, the final display of visual data may contain artefacts. These can be classified into four types based on their origin: capture, processing (coding and decoding), delivery and display. The most significant cause of artefacts is the limited availability of bandwidth, and the tradeoffs associated with using this optimally. Often objective measures of image quality are poorly correlated with perceived image quality. This makes the effectiveness of various mitigation techniques harder to quantify.

/pdf/artefacts-in-image-and-video-systems-classification-and-575rm1lehc.pdf

Artefacts in Image and Video Systems: Classification and Mitigation

In this paper, an extended intelligent edge-based line average-EIELA is proposed. The EIELA could be used in intra-field de-interlacing to remove the jagged edges and the blurring effect of the bilinear interpolation or the traditional algorithms. The EIELA could change its performance dynamically according to the user's requirement. The proposed method can achieve the highest image quality of the traditional ELA if all the taps in the module are enabled, and it also possesses the ability to reduce the taps usage and instruction counts through the turn-off of some calculating sub-modules. The results and VLSI implementation show the proposed method achieve high image quality and low hardware complexity.

/pdf/extended-intelligent-edge-based-line-average-with-its-yv8249m0x5.pdf

Extended intelligent edge-based line average with its implementation and test method

Abstract: The introduction of High definition television (HDTV) asks for spatial up-conversion techniques enabling the display of standard definition material (SDTV). The paper provides an overview of some important linear and non-linear methods. We further include a performance evaluation using a Mean Square Error criterion and show relevant screen shots to allow for a subjective impression obtained with the described algorithms.

Towards an Overview of Spatial Up-conversion Techniques.

Current NTSC systems employ an interlaced scan scheme where visually annoying artifacts, such as line flickers, large area flickers, and edge flickers, are inevitably generated and more objectionable with a larger display. With the development of HDTV systems, how an NTSC signal can be displayed in an HDTV system has also become a problem. Accordingly in order to solve these problems, many adaptive scan rate up-conversion systems have been proposed. The proposed system consists of a modified edge directional line averaging (ELA) interpolation module and contrast enhancement module. The modified ELA interpolation module reduces the visually annoying artifacts induced by an interlaced scan scheme whereas the degradation in the contrast and details due to interpolation processing is compensated by the contrast enhancement module. Moreover, the proposed system is easy to implement and the VLSI architecture is very simple. This new system was mainly developed for doubling the scan rate in an NTSC system; however, it can also be used in the format conversion in an HDTV system.

Sang Um Choi

Papers

Adaptive scan rate up-conversion system based on human visual characteristics