In the camera manufacturing, special methods are needed to detect blemishes occurring on the camera sensor pixels. A blemish is referred as a region of pixels in the camera sensor that are somewhat darker than the background. The blemishes are difficult to detect accurately, but on the other hand, they cause a significant reduction in camera quality. We present a novel filtering method for the blemish detection. The method is based on image scaling, filtering, and difference image calculation that is very fast and accurate in the detection of blemishes. In addition, the algorithm can cope with unprocessed raw image data, in which various distortions, such as noise and vignetting, can be present.

A new and improved visual-oriented grayscale frames interpolation method consists of partially changing, step by step using growing structuring elements, the morphological S transforms of the foreground content of an input frame with the morphological S transforms of the foreground content of an output frame. Better performance comes at the expense of not very great computational complexity. Computer simulations illustrate the results.

Performance of a contemporary two-dimensional face-recognition system has not been satisfied due to the variation in lighting. As a result, many works of solving illumination variation in face recognition have been carried out in past decades. Among them, the Illumination-Reflectance model is one of the generic models that is used to separate the individual reflectance and illumination components of an object. The illumination component can be removed by means of image-processing techniques to regain an intrinsic face feature, which is depicted by the reflectance component. We present a wavelet-based illumination invariant algorithm as a preprocessing technique for face recognition. On the basis of the multiresolution nature of wavelet analysis, we decompose both illumination and reflectance components from a face image in a systematic way. The illumination component wherein resides in the low-spatial-frequency subband can be eliminated efficiently. This technique works out very advantageously for achieving higher recognition performance on YaleB, CMU PIE, and FRGC face databases.

Filter optimization is investigated to design digital camera color filters that achieved high color accuracy and low image noise when a sensor's inherent photon shot noise is considered. In a computer simulation, both RGB- and CMY-type filter sets are examined. Although CMY filters collect more photons, performance is worse than for RGB filters in terms of either color reproduction or noise due to the large noise amplification during the color transformation. When RGB filter sets are used and photon shot noise is considered, the peak wavelength of the R channel should be longer (620 to 630 nm) than the case when only color reproduction is considered: peak wavelengths 600, 550, and 450 nm for RGB channels, respectively. Increasing the wavelength reduces noise fluctuation along the a^* axis, the most prominent noise component in the latter case; however, color accuracy is reduced. The tradeoff between image noise and color accuracy due to the peak wavelength of the R channel leads to a four-channel camera consisting of two R sensors and G and B. One of the two R channels is selected according to the difference in levels to reduce noise while maintaining accurate color reproduction.

We present a novel image compression technique using a classified vector Quantizer and singular value decomposition for the efficient representation of still images. The proposed method is called hybrid classified vector quantization. It involves a simple but efficient classifier-based gradient method in the spatial domain, which employs only one threshold to determine the class of the input image block, and uses three AC coefficients of discrete cosine transform coefficients to determine the orientation of the block without employing any threshold. The proposed technique is benchmarked with each of the standard vector quantizers generated using the k-means algorithm, standard classified vector quantizer schemes, and JPEG-2000. Simulation results indicate that the proposed approach alleviates edge degradation and can reconstruct good visual quality images with higher peak signal-to-noise ratio than the benchmarked techniques, or be competitive with them.

The novel technique of Laplacian eigenmaps (LE) is studied as a means of improving the clustering-based segmentation of color images. Taking advantage of the ability of the LE algorithm to learn the actual manifold of the multivariate data, a computationally efficient scheme is introduced. After embedding the local image characteristics, extracted from overlapping regions, in a high-dimensional feature space, the skeleton of the intrinsically low-dimensional manifold is constructed using spectral graph theory. Using the LE-based dimensionality reduction technique, a low-dimensional map is computed in which the variations of the local image characteristics are presented in the context of global image variation. The nonlinear projections on this map serve as inputs to the Fuzzy C-Means (FCM) algorithm, boosting its clustering performance significantly. The final segmentation is produced by a simple labeling scheme. The application of the presented approach to color images is very encouraging and illustrates the effectiveness of the performance over alternative methods.

To eliminate side-oblique image motion, a fast image algorithm is proposed for implementation on aerial camera systems. When an aerial camera works at a side-oblique angle, much parallel image motion with different rates will occur on the focal plane array simultaneously. Through analysis of how different rates of parallel image motion blur are generated and the relationship between image motion and the field of view (FOV) angle in side-oblique situations, the entire blurred image can be segmented into many slices by their different rates of image motion. To be computed quickly, the slices are divided into pixel lines continuously, and then a specific parallel computing scheme is presented using 1-D Wiener filters to restore all the pixel lines. With all the resulting pixel lines combined, the restoration image comes into being. The experiment results show that the proposed algorithm can effectively restore the details of side-oblique blurred images. The peak signal-to-noise ratio (PSNR) of the restored image can reach 31.426. With the help of the parallel computing capability of a graphics processing unit (GPU), the proposed algorithm can restore a 2048×2048 8-bit blurred image in 17 ms, realizing real-time restoration.

This work introduces a microstereoscopic camera that employs two microvideo sensors and a micromotion stage for automatic vergence and focus control. The parallel-axis stereo configuration using two video sensors enables the stereo camera to converge and focus simultaneously on a real object. To calibrate both vergence and focus with respect to object distance, we derive a nonlinear equation relating image disparity to object distance. Distance measurement using the equation yields very accurate results at near distances. To verify focus control, image focus of different object distances with and without automatic focusing is measured and compared. 3-D reconstruction results of real objects are also presented.

Active contours, as a technique for boundary extraction, have been successfully used in image processing and computer vision. One of the knotty problems of active contours is to conform to the object boundary with complex shape, which could bring heavy manual work at the initialization procedure. The gradient vector flow (GVF) field has been one of the most popular external forces that can increase the capture range of active contours and bidirectionally evolve the active contours toward the object boundary. However, it has a poor performance when dealing with some complex shapes, such as semi-closed concave, screwy concave, hooked concave, as well as the others presented in our experiments. We propose a novel GVF-based balloon force, which can efficiently assist the GVF field in driving active contours toward the complex object shapes. This additional force is used only when the active contours are prevented from evolving toward the object boundary by the saddle and/or stationary points in the GVF field. Therefore, it can maintain the bidirectional evolution property of the GVF and meanwhile take advantage of the power of the balloon force in segmenting complex shapes. Various experimental results on image segmentation are presented to show the good performance of the proposed active contour model that uses the GVF field and the proposed balloon force together.

An autocorrelation function (ACF) to synchronize watermarks has been adopted in practical applications because of its robustness against affine transforms. However, ACFs are vulnerable to projective transform, which commonly occurs during the illegal copying of cinema footage due to the angle of the camcorder relative to the screen. The cinema footage that is captured by camcorders both is projected and has undergone digital-to-analog and analog-to-digital conversion (D-A/A-D conversion). We present a novel watermarking scheme that uses a local autocorrelation function (LACF) that can resist projective transforms as well as affine transforms. A watermark also used for synchronization is designed and additively embedded in the spatial domain. The embedded watermark is extracted in a blind way after recovering from distortions. The LACF scheme with a mathematical model is proposed to synchronize the watermark against distortions. On various video clips, experimental results show that the presented scheme is robust against projective distortions as well as D-A/A-D conversion.

An efficient feature extraction algorithm based on optimized Gabor filters and a relative variation analysis approach is proposed for iris recognition. The Gabor filters are optimized by using the particle swarm algorithm to adjust the parameters. Moreover, a sequential scheme is developed to determine the number of filters in the optimal Gabor filter bank. In the preprocessing step, the lower part of the iris image is unwrapped and normalized to a rectangular block that is then decomposed by the optimal Gabor filters. After that, a simple encoding method is adopted to generate a compact iris code. Experimental results show that with a smaller iris code size, the proposed method can produce comparable performance to that of the existing iris recognition systems.

This study reports about the detection of non-natural structures in outdoor natural scenes. In particular, we present a new approach based on ridgelet transform for the segmentation of man-made objects in landscape scenes. Multiscale directional moments of ridgelet coefficients are used as features along with a principal component analysis (PCA) followed by a linear discriminant analysis (LDA), kernel-based LDA (KLDA), or support vector classifier (SVC). The statistical learning is done on about 3,000 image patches that represent natural and artificial content. Performances are measured in terms of image patch type classification (natural versus non-natural) and man-made object segmentation on two different image test sets. Results using ridgelets are compared to Gabor features. Altogether, we compare performance for six different feature/classifier combinations: ridgelets+LDA, ridgelet+KLDA, ridgelets+SVC, Gabor+LDA, Gabor+KLDA, and Gabor+SVC, and various external parameter values. Results show that most of the time, the combinations with ridgelets provide comparable or better performance.

We present a modification of the standard time-frequency (t-f) analysis for landmine detection. The modification is adapted to synthetic aperture radar (SAR) images that may or may not exhibit near-circular symmetry. Each prospective SAR image is here sliced along several directions to generate t-f plots along those chosen cuts, and the resulting 2-D plots are correlated in pairs to obtain a relevant metric, which is defined as their ratio. This metric has served to distinguish targets from clutter objects in the various cases examined here. The procedure was validated using a dataset obtained in a recent field test, and the results are shown.

We propose a new method for tamper localization and restoration using noise pixels in binary document images. For such images, it is difficult to find a sufficient number of low-distortion pixels in individual blocks with blind detection property. Also, a perceptual watermark cannot be embedded in white regions of the document image, making such regions insecure against hostile attacks. An erasable watermark is embedded in each block of the document image independently. The embedding process introduces some background noise. However, the content in the document can be interpreted by the user, because human vision has the inherent capability to recognize various patterns in the presence of noise. If authenticity is verified for the content of each block, the exact copy of original image is restored at the blind detector for further use and analysis. Experimental results show that an erasable watermark of necessary data length can be embedded in individual blocks to attain effective localization and restoration capability. Using the proposed method, it is possible to restore the original text sequence in text document images after multiple alterations like text deletion, insertion, substitution, and block swapping.

For many image processing applications, edge detection is a very important task that needs to be assessed, since the success or failure of these applications depends on the performance of this task. Assessment of edge detection is largely subjective; however, current trends in the image processing community are moving toward objective assessment. In recent years, many different methods have been proposed to assess edge detection, although no agreement has been reached as the proper method, since previous comparisons have produced contrasting results. A comparison of assessment methods using an objective approach is presented. Methods are compared by analyzing the results of an optimization procedure using genetic algorithms and the methods as fitness. The comparison is based on the premise that better assessment methods will lead the optimization procedure to produce better results. A cross-validation is carried out to compare the results obtained using one assessment method with others. Conclusions provide recommendations for authors interested in assessing edge detection algorithms.

This PDF file contains the errata for “JEI Vol. 18 Issue 02 Paper JEI ERR-APR2009-1” for JEI Vol. 18 Issue 02

This PDF file contains the errata for “JEI Vol. 18 Issue 02 Paper JEI ERR-APR2009-2” for JEI Vol. 18 Issue 02

This PDF file contains the editorial “Machine Learning for Audio, Image and Video Analysis” for JEI Vol. 18 Issue 02