عنوان مقاله
کانتور فعال: شیوه الگوریتم ژنتیکی موازی
فهرست مطالب
مقدمه
کانتور فعال
بهینه سازی الگوریتم ژنتیک
محاسبه موازی
نتیجه گیری
بخشی از مقاله
• کراس اور: یک جفت پدر برای تولید راه حل فرزند انتخاب می شوند. نسل جدید بسیاری از ویژگیهای والدینشان(پدرانشان) را دارا می باشند.
• جهش: راه حل پدر برای تولید راه حل فرزند انتخاب می شود. نسل جدیداز اطلاعات پدرانشان با تغییرات و اصلاحات تصادفی سهم می برد( دارای اطلاعات مشترک میباشد).
مراحل قبل تا زمان دستیابی به شرایط خاتمه دادن و پایان دهی، تکرار می شوند: راه حلی یافت می شود که معیارهای مینیموم، تعداد ثابت نسل های حاصل شده، نیل به بودجه تخصیص داده شده و غیره را رعایت و تامین می کند.
کلمات کلیدی:
Active contour: a parallel genetic algorithm approach Florence Kussener1 1 MathWorks, 2 rue de Paris 92196 Meudon Cedex, France Florence.Kussener@mathworks.fr Abstract This paper presents an algorithm for automatically detecting contours using snake algorithm. Prior knowledge is first used to locate initial contours for the snakes. Next we optimize the energy by using genetic algorithm approach. We introduce parallel computing to reduce computation time for the genetic algorithm calculations. Key words Active contour, parallel computing, genetic algorithm, segmentation, snake 1 Introduction Edge detection is a fundamental tool in image processing, image pattern recognition, and computer vision techniques. Ideally, the result of applying an edge detector to an image should lead to a set of curves that indicate the contour of objects as well as curves that correspond to discontinuities in surface orientation. Furthermore, applying an edge detection algorithm to an image may reduce the amount of data to be processed while preserving the structural properties of an image. Edge detection is not a trivial task. There are many methods for edge detection. In this paper, we will introduce the active contour method, or snake algorithm, which minimizes the energy function. Active contours have multiple advantages over classical feature attraction techniques. Snakes are easy to manipulate using external image forces. They are self-adapting in their search for a minimal energy state. Furthermore they can be used to track dynamic objects in temporal as well as the spatial dimensions. Nevertheless, one of the biggest drawbacks of this method is that snakes are get stuck in local minima states. This may be overcome by using genetic algorithm techniques at the expense of longer computation times. Genetic Algorithm is an optimization solver, which does an analogy to Darwin evolution by combining mutation, crossover and selection step. One of the biggest advantages of Genetic Algorithm is its ability to find a global optimum. A consequence of this search is to have an additional computation time. This is why Parallel Computing is a good approach to optimize time of computation.
