Notification:
We are currently experiencing intermittent issues impacting performance. We apologize for the inconvenience.
By Topic

Image Processing, IET

Issue 2 • Date March 2012

Filter Results

Displaying Results 1 - 10 of 10
  • Precise multi-level face detector for advanced analysis of facial images

    Publication Year: 2012 , Page(s): 95 - 103
    Cited by:  Papers (3)
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (616 KB)  

    This study presents a new method for precise detection of frontal human faces and eyes using a multi-level ellipse detector combined with a support vector machines verifier. Main contribution of this study lies in improving the accuracy of eye detection in high-quality images, which is often neglected by alternative methods. Although many approaches to face detection have been proposed recently, relatively little attention has been paid to the detection precision. It is worth noting that the detection precision is particularly important for face analysis purposes. More specifically, the authors demonstrate that the detection error propagation substantially affects the face recognition performance. With the proposed improvements the authors have managed to increase the face recognition rate by 7.7% for AR database compared with the publicly-available implementation of the well-established Viola-Jones face and eye detector. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Very-large-scale integration design of a low-power and cost-effective context-based adaptive variable length coding decoder for H.264/AVC portable applications

    Publication Year: 2012 , Page(s): 104 - 114
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (1053 KB)  

    Context-based adaptive variable length coding (CAVLC) is a new and efficient entropy coding tool in H.264/AVC (advanced video coding). This study proposes a low-power and cost-effective CAVLC decoding architecture for the H.264/AVC baseline profile. Specifically, this study proposes an optimum two-layer power model for the variable length look-up table (VLUT) in the CAVLC decoder, and divides the decoding phase of the LUT into two-layer decoding to reduce power consumption. To achieve a cost-effective design, the proposed design merges common codewords to reduce the hardware cost among different LUTs in the second layer decoding. The proposed decoder is based on Taiwan Semiconductor Manufacturing Company (TSMC) 0.18 μm CMOS technology, and was completely verified on a field-programmable gate array (FPGA) emulation platform. The proposed design meets the demands of the real-time CAVLC decoding and reduces power consumption by 44-48% more than previous low-power CAVLD schemes. Finally, the proposed low-power and cost-effective CAVLD design is suitable for H.264/AVC portable applications. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Visual tracking via bag of features

    Publication Year: 2012 , Page(s): 115 - 128
    Cited by:  Papers (5)
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (1621 KB)  

    In this paper, we propose a visual tracking approach based on `bag of features` (BoF) algorithm. First we use incremental PCA visual tracking (IVT) in the first few frames and collect image patches randomly sampled within the tracked object region in each frame for constructing the codebook; the tracked object then can be converted to a bag. Second we construct two codebooks using color (RGB) features and local binary pattern (LBP) features instead of only one codebook in traditional BoF, thereby extracting more informative details. We also devise an updating mechanism to deal with pose and appearance changes of objects. In the tracking process, a constant number of candidates are generated by sampling technique in each frame. Image patches are then randomly sampled and candidates are represented as bags by codebooks. Thus, we can compute patch similarity of a candidate with the codewords and bag similarity with trained bags. The actual object is then located by finding the maximal combined similarity of patches and bags. Experiments demonstrate that our approach is robust in handling occlusion, scaling and rotation. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Colour correction pre-processing and chrominance reconstruction post-processing for multi-view video coding

    Publication Year: 2012 , Page(s): 129 - 138
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (1062 KB)  

    In a multi-view video system, a great amount of data for storage and transmission raise an inevitable problem to be solved. In this study, a new multi-view video coding (MVC) method is proposed by performing colour correction as a pre-processing step and chrominance reconstruction as a post-processing step. In the proposed method, colour correction is first performed to eliminate colour differences among views and to improve coding efficiency simultaneously. Then, chrominance information in some particular views is discarded to reduce bit-rates. Finally, the discarded chrominance information is reconstructed accurately at the decoder. Experimental results show that the proposed method can better increase the bit-rate saving than the original MVC system. Moreover, the proposed method can accomplish higher quality reconstruction at the trivial cost of quality degradation. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Padua point interpolation and Lp-norm minimisation in colour-based image indexing and retrieval

    Publication Year: 2012 , Page(s): 139 - 147
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (242 KB)  

    Colour has proven to be a very powerful feature for image indexing. Many examples of image retrieval systems based on colour or chromaticity histograms have been proposed, following on from the histogram intersection method of Swain and Ballard. Here the authors introduce a compact representation of the chromaticity histogram which is based on the Padua point interpolation technique. Specifically, the histogram is represented as a linear combination of Chebyshev polynomials. This bounds a certain maximum deviation, as opposed to a least-squares criterion used in previous work. With this in mind, the minimisation of different Lp norms and the L norm of the error is compared. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Shadow identification for digital imagery using colour and texture cues

    Publication Year: 2012 , Page(s): 148 - 159
    Cited by:  Papers (2)
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (1099 KB)  

    Shadows cause a significant problem for automated systems which attempt to understand scenes, since shadow boundaries may be incorrectly recognised as a material change, and incorrectly recognised as an object. Shadow identification is therefore an important pre-processing step for applications such as shadow removal, shadow invariant object recognition and shadow invariant object tracking. Many existing shadow identification methods are often limited by the types of shadow boundaries (penumbra) which can be found, by the density (darkness) of the shadows and by the type of surface texture on which the shadows are cast. In addition many of these methods are limited to a specific type of scene, while others result in high levels of false positive (FP) shadow identification. To address these problems, a novel algorithm for automatic shadow identification is proposed, which makes use of a new tree-structured segmentation algorithm for candidate shadow region identification, as well as a combination of colour illumination invariance and texture analysis for shadow verification. The method is tested on a number of indoor and outdoor images exhibiting different types of shadows, surfaces and illumination sources. These results indicate that the proposed method performs well against the state of the art; in particular, the rate of FP identification is reduced from 26 to below 13% when compared with using illumination invariance alone. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Spectral technique to recognise occluded objects

    Publication Year: 2012 , Page(s): 160 - 167
    Cited by:  Papers (1)
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (671 KB)  

    When recognising partially visible objects in a scene, a good global decision should be made based on locally gathered features for their recognition, since global information is no longer reliable. This local-to-global nature of occlusion recognition leads us to spectral matching technique. Unfortunately, conventional spectral matching is not desirable for noisy data set from a cluttered scene. In this study, a top-down procedure is introduced into a standard spectral matching for the recognition of occluded objects. Feature points are firstly evaluated and associated with object(s) of interest. Subsequently, geometrical consistency is enforced to find correct correspondences among the candidate matches with high association scores. Based on this two-stage strategy, both appearance and geometric information are taken into consideration. Our algorithm is implemented for both 2D and 3D occluded objects recognition under different occlusion rates. It is shown that the improvement has been made for spectral correspondence algorithm to recognise occluded objects and it has a comparable recognition rate with the state-of-the-art recognition methods. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Colour image super-resolution using geometric grouplets

    Publication Year: 2012 , Page(s): 168 - 180
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (2126 KB)  

    In this work, a technique for generating a super-resolution (SR) image from a single multi-valued low-resolution (LR) input image is proposed. This problem is approached from the perspective of image geometry-oriented interpolation. First, the geometry of the LR image is obtained by computing the grouplet transform. These grouplet bases are used to define a grouplet-based structure tensor to capture the geometry and directional features of the LR colour image. Then, the SR image is synthesised by an adaptive directional interpolation using the extracted geometric information to preserve the sharpness of edges and textures. This is accomplished by the minimisation of a functional, which is defined on the extracted geometric parameters of the LR image and oriented by the geometric flow defined by the grouplet transform. The proposed SR algorithm outperforms the state-of-the-art methods in terms of visual quality of the interpolated image. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Video post processing: low-latency spatiotemporal approach for detection and removal of rain

    Publication Year: 2012 , Page(s): 181 - 196
    Cited by:  Papers (1)
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (1892 KB)  

    In this study, a novel, efficient and simple algorithm for detection and removal of rain from video using spatiotemporal properties is proposed. Advantageously, the spatiotemporal properties are involved to separate rain pixels from non-rain pixels. It is thus possible by way of the proposed algorithm to involve less number of consecutive frames, reducing the buffer size and delay. It works only on the intensity plane which further reduces the complexity and execution time significantly. This new algorithm does not assume the shape, size and velocity of raindrops which makes it robust to different rain conditions. Proposed method reduces the buffer size, which reduces the system cost, delay and power consumption. For performance evaluation, in addition to miss & false detection a new metric spatiotemporal variance is introduced. Results show that the proposed algorithm outperforms the other rain removal algorithms. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Efficient computation of correlation coefficient using negative reference in template matching applications

    Publication Year: 2012 , Page(s): 197 - 204
    Save to Project icon | Click to expandQuick Abstract | PDF file iconPDF (584 KB)  

    Design and implementation of a new early termination algorithm for efficient calculation of correlation coefficient for template matching is presented. The proposed algorithm correlates the candidates against `negative` (bit-inverted) version of reference instead of the original to implement a low cost early termination criterion. Around 10% computational savings have been demonstrated while using the proposed algorithm to compute normalised correlation as error metric for motion estimation in software implementation of H.264 video encoder. The algorithm lends itself to efficient hardware implementation because of its simple cost function. Further hardware savings have been realised by noting that the multiplication products generated by the proposed negative reference correlation algorithm tend to have low magnitudes with significantly less variance than those generated by other schemes. This allows a low-precision summation stage to accumulate majority of the multiplication products without losing the precision of results. Results of the hardware implementation on Xilinx Virtex-5 FPGA have been provided. The overall design is shown to consume around 85% less logic resources and operate at 140% higher speed than existing architectures. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.

Aims & Scope

The range of topics covered by IET Image Processing includes areas related to the generation, processing and communication of visual information.

Full Aims & Scope

Meet Our Editors

Publisher
IET Research Journals
iet_ipr@theiet.org