The compression efficiency of an indexed image depends on the palette ordering. However, obtaining an optimal palette ordering is not an easy task, and approximate solutions are usually sought in practice. Among the available methods, the pairwise merging technique proposed by Memon is the best in terms of compression efficiency. However, this method needs extensive computations. In this letter, a novel algorithm is presented to find an approximate solution of the palette reordering problem. Combined with Pinho's method, our approach is better than the pairwise merging technique in terms of compression efficiency and computing time

In this paper, a new and fast-searching algorithm for vector quantization is presented. Two inequalities, one used for terminating the searching process and the other used to delete impossible codewords, are presented to reduce the distortion computations. Our algorithm makes use of a vector's features (mean value, edge strength, and texture strength) to reject many unlikely codewords that cannot be rejected by other available approaches. Experimental results show that our algorithm is superior to other algorithms in terms of computing time and the number of distortion calculations. Compared with available approaches, our method can reduce the computing time and the number of distortion computations significantly. Compared with the best method of reducing distortion computation, our algorithm can further reduce the number of distortion calculations by 29% to 58.4%. Compared with the best encoding algorithm for vector quantization, our approach also further reduces the computing time by 8% to 47.7%.

To reduce the artifacts due to color filter array (CFA) interpolation, a modified mean-removed classified vector quantization algorithm is proposed. The algorithm extends and modifies vector quantization to discover the relationships between the G channels of interpolated images and their corresponding original versions using the information from CFA images. The discovered relationships are stored in two codebooks and are used to improve the edge and texture quality of G channels of interpolated images. After the interpolated G values are refined, the interpolated R and B values can also be improved using the refined G values. The experimental results show that the proposed algorithm can effectively reduce the artifacts of interpolated CFA images. Comparing our method and the best CFA interpolation algorithm as far as we know, the PSNR improvements of R, G, B channels are 0.89 dB, 0.71 dB, and 0.74 dB, respectively.

Color printers usually have only four inks (Cyan, Magenta, Yellow, and Black). They can not represent all colors of a color image. Traditional error-diffusion algorithm is proposed for solving this problem. In general, error-diffusion produces better halftoned images. Unfortunately, it also produces some artifacts, i.e. unstable phenomenon and smeared edges. These situations are much worse when dot overlap is included. We use a series of methods to improve the quality of halftoned color images displayed by printers with dot overlap. First, we show that the traditional error-diffusion algorithm may become stable if a proper formula is used to determine the color difference and propose a stable error-diffusion algorithm that does not consider the dot overlap. Then, we modify this error-diffusion algorithm as a new one which includes dot overlap. We also present a dot-size estimation method for this algorithm. Finally, the edge enhancement function is added into our error-diffusion algorithm Halftoning a color image by our proposed error-diffusion algorithm with edge enhancement has the following characteristics: the unstable phenomenon does not exit; the color-error caused by dot overlap is corrected and the smeared edges are sharpened.

This correspondence extends and modifies classified vector quantization (CVQ) to solve the problem of inverse halftoning. The proposed process consists of two phases: the encoding phase and decoding phase. The encoding procedure needs a codebook for the encoder which transforms a halftoned image to a set of codeword-indices. The decoding process also requires a different codebook for the decoder which reconstructs a gray-scale image from a set of codeword-indices. Using CVQ, the reconstructed gray-scale image is stored in compressed form and no further compression may be required. This is different from the existing algorithms, which reconstructed a halftoned image in an uncompressed form. The bit rate of encoding a reconstructed image is about 0.51 b/pixel.

This correspondence shows that line and point correspondences can obtain orientation with the same accuracy, and that when using line correspondence, using more matching lines does not guarantee that the orientation error can be reduced. Orientation detection using line correspondence is very sensitive to errors in image location. We provide a sensitivity analysis of line correspondence, and conditions which lead to orientation sensitivity due to errors in image location. The orientation error using line correspondence due to offset from the image center is obtained analytically. A method for selecting three matching lines, such that orientation is not sensitive to image location errors and using more matching edges to improve the accuracy of orientation is given. Finally three examples are given to illustrate the concepts and support the theoretical results. The first example shows that an orientation with reasonable accuracy may be obtained by using three matching lines. The second example demonstrates that both point and line correspondences can obtain orientation with the same accuracy. The last example supports experimentally the result regarding the orientation error using line correspondence due to offset of the image center and lens distortion.>

In this paper, a generic geometrical path planning is developed for the coordinated motion of two manipulators. For a redundant robotic system consisting of two non-redundant manipulators, we can define two trajectories for these two manipulators, by applying the method presented in this paper, and use the direct inverse kinematic solution of each manipulator. Thus the intrinsic inaccuracy problem of the pseudo-inverse, which the conventional method had suffered, may be avoided. If the trajectory lT, for the first manipulator, is not immaterial, then the coordinated motion is planned to reduce the execution time. If the defined path lT should be followed by the first manipulator, then the second trajectory 2T is scheduled according to the given task for the coordinated motion. It is believed that the method, presented here, is the first of its kind to plan two separate trajectories, in general case, for two manipulators during the coordinated motion.

Since occlusion is present in all but the most constrained environments, the recognition of partially occluded objects is of prime importance. However, almost all of the existing algorithms which resolve the occlusion problem are complicated and required extensive computation. An attempt is thus made to develop a recognition system with moderate cost and speed, and high rate of recognition. The authors present a hierarchical problem-solving technique to match models to objects in the image. They evaluate the system and find that the storage requirements are less than those of other systems. Furthermore, the time complexity of the system is merely a linear function of the number of feature points.>

An efficient algorithm is developed for the motion control of manipulators in the neighborhood of wrist singularity. The method, based on a scheduling scheme, guarantees a robot's location accuracy, maintains a desired Cartesian motion speed, and does not involve the complex computations of a robot's Jacobian and its inverse. In case studies, three algorithms, namely the maximum joint rates, the iterative method, and the proposed scheduling method, are compared. The scheduling scheme generally gives more satisfactory results in terms of location and orientation accuracies. It is believed that the combination of simple computation, acceptable orientation error, and the ability of maintaining desired speed should make the proposed method very attractive in practical application.>

The inverse solutions of a robotic systems are generally produced by a serial process. Due to the computing time of processing geometry data and generating an inverse solution corresponding to a specified point in Cartesian trajectory is larger than the sampling period, the missing points in the joint space are generated by some interpolation schemes (linear or cubic spline interpolations) between two inverse solutions. The dynamic errors are therefore introduced. Obviously this kind of dynamic errors can be eliminated, if the computational time of generating the inverse solutions and processing geometric information can be less than the sampling period. The dynamic errors are increased when a robot's speed is increased by using the available schemes. For a high speed and high performance robot, the dynamic errors can be significant. In this paper, a parallel algorithm for a robot's inverse kinematics is derived and corresponding VLSI architectures are presented. This algorithm can also be implemented by using multiprocessors. By using the proposed parallel algorithm, it is believed that the dynamic errors can be reduced significantly or even eliminated if the computing time of processing geometric data can be reduced significantly by using the technique of parallel processing.