Acil Çeviri Lazım.. (2 sayfa)

[godzilla112]

Saygıdeğer pati amcalar bu geç saatte rahatsız ettim çok özür diliyorum. Lütfen bana bi yardım eli uzatın bu yazıyı Türkçeye çevirin. Valla çevirin, üzerinizden hayır duamı eksik etmiycem. (ingilizcem kısıtlı, bugün olduğumuz trigonometri yazılısı ve çevirdiğim bir kaç sayfa sınırları iyice daralttı bu yüzden yardıma muhtacım). Yarın sabaha (07:20) kadar yetiştirmem lazım...

The following sections describe the developed method for structural matching. Some application examples involving relative orientation of stereo image pairs, image matching of radar images and tie point measurement in aerotriangulation are illustrated. They show that with the application of structural matching digital photogrammetry can reach a very high level of automation. The “black box” philosophy for photogeometric operations, which has been predicted by some photogeometric experts (e.g. Ackermann, 1991), is further realized in this contribution.

An approach for structural matching

2.1 Structural description
Structural matching establishes a correspondence between two structural descriptions. A structural description of an image consists of a set of image features and the relationships among the features as Fig.1 illustrates. Features ( points, lines and regions) are also referred to as primitives. Each primitive and relation can be described by several attributes. For example a point primitive pi can be described as:

The goal of structural matching is to find a correspondence of the best match between the primitives and relations of two structural descriptions. The available primitives and relations of an image can be enormous Two images are usually only partly overlapping, so the correct match of two descriptions may be only a partial correspondence of all primitives and relations. One primitive (or relation) of the first description. Therefore, there may exist many possible matches between two structural descriptions. The search time for the best matches can be very long. In order to solve the structural matching problem quickly and correctly, topics like evaluation function, search method, correctness check and the efficent extraction of structural descriptions must be researched and resolved.

2.2 Evaluation Function
Suppose that two images are to be matched and Dl is used to represend a structural description of the second image According to the maximum-likelihood estimation the best match of two descriptions hb should have the maximal conditional probability among all possible matches h1,h2…hn i.e:

As Eq. 4 shows, maximizing P(hi/Dl, Dr) is equal to maximizing P(Dl,Dr/hi). Therefore, the coint probability P(Dl, Dr /hi) can be taken as an evaluation criterion for the goodness of a possible match. Thus,

Where subscripts j and k represent the primitibes of the descriptions, In practice, the joint probability P(Dl, Dr / hi)can be calculated according to the similarity of the attributes of primitives and relations. In most cases a different weight is used for each attribute.

2.3 Search Method

Tree search methods are often used to find solutions for many artificial inteligence problems (Nilsson, 1982). Tree search methods can be divided into two types: blind search methods and informed search methods. The blind search method treats all the tree nodes equally, so they usually take a long time to find solution. The informed search method uses some measures to evaluate the nodes and first try the node which is more probable to lead to the solution. An investigation of the search methods can be found in Vosselman (1994).
An informed search method is also employed in tihs research. It utilies the similarity of the tree nodes to guide the search. In order to reduce the search time, which increases exponentially with the number of primitives of the structural descriptions, some further measures are developed and integrated into the search method. They include the substructure concept, primitive ordering best minimum patch and geometrical constraints.
With the substructure description are integrated based on their connectibity so that the primitive volume can be decreased significantly. In the case of structural matching of two images, the structural description can be reorganized as idicated in Fig. 2. This is possible because in an image the points, lines and regions are interconnected with certain relatiıons and the attributes of a point primitive can be calculated mor accurately and reliably than those of a line or region. Each point with its associated lines and regions forms a substructure. The search for matched primitives becomes a search for matched substructures. Thus, the search space is reduced significantly.
In primitive ordering the primitives or the primitive groups are reordered according to their type, weights and similarity, so that the primitives, which more likely lead to the solution, are tried earlier than the others by the search method.[signature][hline][small

[godzilla112]

Ya beyler lütfen bu gerçekten benim için önemli yarın sabaha kadar çeviremezsem öldüm bittim ben. mesaj 26 kere okunmuş ama bunlardan 20si zaten benim. Gecenin bu yarısı kimi bulursam kim bu mesaja bakarsa ondan istiyorum lütfen..
Bu yazıyı sabaha kadar yırtınsam çeviremem.. Bide çevirmeye başlayanlar bana ö.m atarsa -ben çeviriyorum sen kendini yırtma, üzülme gariban insan diye- bende kendimi yırtmayı bırakırım.. Yinede teşekkürler..[signature][hline]
Mana alassë equë inyë Turk!!!
K.S.D.M.P.A.L

[Bu mesaj godzilla112 tarafından 20 Kasım 2004 01:55 tarihinde değiştirilmiştir]