Variable Search Space Converging Genetic Algorithm for Solving System of Non-linear Equations

This paper introduce a new variant of the Genetic Algorithm whichis developed to handle multivariable, multi-objective and very high search space optimization problems like the solving system of non-linear equations. It is an integer coded Genetic Algorithm with conventional cross over and mutation but with Inverse algorithm is varying its search space by varying its digit length on every cycle and it does a fine search followed by a coarse search. And its solution to the optimization problem will converge to precise value over the cycles. Every equation of the system is considered as a single minimization objective function. Multiple objectives are converted to a single fitness function by summing their absolute values. Some difficult test functions for optimization and applications are used to evaluate this algorithm. The results prove that this algorithm is capable to produce promising and precise results.

Automatic Histogram Specification for Glioma Grading Using Multicenter Data

Multicenter sharing is an effective method to increase the data size for glioma research, but the data inconsistency among different institutions hindered the efficiency. This paper proposes a histogram specification with automatic selection of reference frames for magnetic resonance images to alleviate this problem (HSASR). The selection of reference frames is automatically performed by an optimized grid search strategy with coarse and fine search. The search range is firstly narrowed by coarse search of intraglioma samples, and then the suitable reference frame in histogram is selected by fine search within the sample selected by coarse search. Validation experiments are conducted on two datasets GliomaHPPH2018 and BraTS2017 to perform glioma grading. The results demonstrate the high performance of the proposed method. On the mixed dataset, the average AUC, accuracy, sensitivity, and specificity are 0.9786, 94.13%, 94.64%, and 93.00%, respectively. It is about 15% higher on all indicators compared with those without HSASR and has a slight advantage over the result of a manually selected reference frame by radiologists. Results show that our methods can effectively alleviate multicenter data inconsistencies and lift the performance of the prediction model.

Reverse Search Particle Swarm Optimization to Deal with Dynamic Optimization Problems

Multiswarm approaches are used in many literatures to deal with dynamic optimization problems (DOPs). Each swarm tries to find promising areas where usually peaks lie and many good results have been obtained. However, steep peaks are difficult to be found with multiswarm approaches , which hinders the performance of the algorithm to be improved furtherly. Aiming at the bottleneck, the paper introduces the idea of sequential niche technique to traditional multiswarm approach and thus proposes a novel algorithm called reverse space search multiswarm particle swarm optimization (RSPSO) for DOPs. RSPSO uses the information of the peaks found by coarse search of traditional multiswarm approach to modify the original fitness function. A newly generated subswarm - reverse search subswarm evolves with the modified fitness function, at the same time, other subswarms using traditional mltiswarm approach still evolve. Two kinds of subswarm evolve in cooperation. Reverse search subswarm tends to find much steeper peak and so more promising area where peaks lie is explored. Elaborated experiments on MPB show the introduction of reverse search enhances the ability of finding peaks , the performance of RSPSO significantly outperforms traditional multiswarm approaches and it has better robustness to adapt to dynamic environment with wider-range change severity.

Improvement to Multi-resolution Collective Detection in GNSS Receivers

Collective detection is a promising approach to positioning in a weak signal environment, in which the navigation solution is directly obtained by acquisition search in a multi-dimensional position and common clock bias uncertainty space. By combining the correlation values from multiple satellites and fully utilizing the coherence between them, the detectable C/N0 of individual satellites can be lowered. However, the lack of a computationally efficient optimization algorithm due to high dimensionality and complexity has hindered its application. A multi-resolution collective detection is therefore proposed to be a coarse-to-fine searching approach to solve for the position and common clock bias estimation. Although it reduces the computation time of collective detection, there is a gap in the efficiency study, which is the contribution of this research. The features of different levels of search in a multi-resolution algorithm are investigated. For a coarse search with large horizontal position step size, a smaller common clock bias step size is proposed instead of an averaging correlogram to reduce computation complexity as well as to obtain high time resolution. For the fine search with small horizontal space step size, a 3-D Dichotomous searching scheme is designed and applied to reduce the number of searching grids. Computer simulation results using experimental raw data are provided, to demonstrate the performance improvement against the conventional methods.

Base on SIFT-Harris Operator of the Document Image Matching Method

Abstract：Proposed a method of document image matching based on SIFT-Harris operator, Use of SIFT-Harris operator to accurately search match for the same name point. First use SIFT operator for the coarse search match to find out a rough affine transformation relationship of matched Image and based image, Then used the Harris operator and gray correlation matching algorithm refined search. It improved the matching speed and accuracy. Experimental results show that the method works well for document image matching.

A Novel Global Optimization Method – Genetic Pattern Search

A novel global optimization method is proposed to find global minimal points more effectively and quickly. The new algorithm is based on both genetic algorithms (GA) and pattern search (PS) algorithms, thus, we have named it genetic pattern search. The procedure involves two-phases: First, GA executes a coarse search, PS then executes a fine search. Experiments on four different test functions (consisting of Hump, Powell, Rosenbrock, and Woods) demonstrate that this proposed new algorithm is superior to improved GA and improved PS with respect to success rate and computation time. Therefore, genetic pattern search is an effective and viable global optimization method.

AUTOMATIC EXTRACTION OF CHARACTERS IN COMPLEX SCENE IMAGES

We have developed a generalized alphanumeric character extraction algorithm that can efficiently and accurately locate and extract characters from complex scene images. A scene image may be complex due to the following reasons: (1) the characters are embedded in an image with other objects, such as structural bars, company logos and smears; (2) the characters may be painted or printed in any color including white, and the background color may differ only slightly from that of the characters; (3) the font, size and format of the characters may be different; and (4) the lighting may be uneven. The main contribution of this research is that it permits the quick and accurate extraction of characters in a complex scene. A coarse search technique is used to locate potential characters, and then a fine grouping technique is used to extract characters accurately. Several additional techniques in the postprocessing phase eliminate spurious as well as overlapping characters. Experimental results of segmenting characters written on cargo container surfaces show that the system is feasible under real-life constraints. The program has been installed as part of a vision system which verifies container codes on vehicles passing through the Port of Singapore.