كليدواژه :
پايگاه داده توصيفي مكان محور , بناهاي تاريخي , تابع پالس پويا , بافت هاي تصويري تصادفي
چكيده فارسي :
تابع پالس پويا (DPF) (Dynamic Pulse Function) نسخه بهبوديافته‌اي از (PF (Pulse Function است. اين تابع تصوير نهايي را بر اساس هندسه واقعي و متناسب با عرض و ارتفاع نما توليد مي نمايد. الگوريتم DPF در زمان نگاشت آن بر روي مدل سه‌بعدي، از اعوجاج تصوير جلوگيري مي كند. با توجه به اينكه هر دو الگوريتم تابع پالس و تابع پالس پويا نوعي از روش‌هاي بافت سازي رويه اي محسوب مي‌شوند، در اين مقاله هدف ايجاد مدل هايي با سطح جزييات رده 3 (LoD3) است. در روش‌هاي برگرفته از الگوريتم DPF كه با جزئيات LoD2 ايجاد مي‌شوند، هندسه پنجره ها و درها در يك فايل شماي (XML (Extensible Markup Language كه هيچ اتصالي با مدل سه‌بعدي براي مثال در فرمت CityGML (فرمت تبديل داده-Data exchange format) ندارد، ذخيره مي‌گردد و بافت تصويري به شكل دوبعدي و با واقع‌گرايي بالايي بدون مشكل خوابيدگي هندسه هاي فرورفته يا برآمده، ايجاد مي‌گردد. در اين تحقيق، پارامترهاي مربوط به DPF با به‌كارگيري زبان برنامه‌نويسي Ruby در نرم‌افزار SketchUp Trimble، به‌منظور توليد خودكار موقعيت دقيق و عمق پنجره‌ها و درها با جزئيات LoD3 و بر اساس الگوريتم DPF به‌دست‌آمده است. مزيت استفاده از اين روش، توليد خودكار تعداد انبوهي از هندسي هاي مشابه و بافت‌هاي تصادفي از نماي واقعي ساختمان است. پنجره ها و درها و ساير المان‌هاي نما با استفاده از پارامترهاي الهام گرفته از DPF همراه با تعريف موجوديت ها و لايه هاي پنجره ها و عمق دقيق آن‌ها، به‌طور خودكار ايجاد مي‌گردند. اين الگوريتم مي‌تواند كاربرد بسياري در مدلسازي بناهاي تاريخي داشته باشد كه علاوه بر كيفيت بالا و حجم پايين، در آن مي توان پايگاه داده مكان محور توصيفي براي همه المان‌ها ايجاد كرد.
چكيده لاتين :
The pulse function (PF) is a technique based on procedural preprocessing system to generate a
computerized virtual photo of the façade with in a fixed square size. Dynamic Pulse Function
(DPF) is an enhanced version of PF which can create the final photo, proportional to real
geometry. This can avoid distortion while projecting the computerized photo on the generated 3D
model. The challenging issue that might be handled for having 3D model in LoD3 rather than
LOD2, is the final aim that have been achieved in this paper. In this research the parameters of
Dynamic Pulse Functions are utilized via Ruby programming language in SketchUp Trimble to
generate (exact position and deepness) the windows and doors automatically in LoD3 based on the
same concept of DPF. The advantage of this technique is automatic generation of huge number of
similar geometries e.g. windows by utilizing parameters of DPF along with defining entities and
window layers. In case of converting the SKP file to CityGML via FME software or CityGML
plugins the 3D model contains the semantic database about the entities and window layers which
can connect the CityGML to MySQL. The concept behind DPF, is to use logical operations to
project the texture on the background image which is dynamically proportional to real geometry.
The process of projection is based on two vertical and horizontal dynamic pulses starting from
upper-left corner of the background wall in down and right directions respectively based on image
coordinate system. The logical one/zero on the intersections of two vertical and horizontal
dynamic pulses projects/does not project the texture on the background image. It is possible to
define priority for each layer. For instance the priority of the door layer can be higher than window
layer which means that window texture cannot be projected on the door layer. Orthogonal and
rectified perpendicular symmetric photos of the 3D objects that are proportional to the real façade
geometry must be utilized for the generation of the output frame for DPF. The DPF produces very
high quality and small data size of output image files in quite smaller dimension compare with the
photorealistic texturing method. The disadvantage of DPF is its preprocessing method to generate
output image file rather than online processing to generate the texture within the 3D environment
such as CityGML. Furthermore the result of DPF can be utilized for 3D model in LOD2 rather
than LOD3. In the current work the random textures of the window layers are created based on
parameters of DPF within Ruby console of SketchUp Trimble to generate the deeper geometries of
the windows and their exact position on the façade automatically along with random textures to
increase Level of Realism (LoR). As the output frame in DPF is proportional to real geometry
(height and width of the façade) it is possible to query the XML database and convert them to units
such as meter automatically. In this technique, the perpendicular terrestrial photo from the façade is
rectified by employing projective transformation based on the frame which is in constrain
proportion to real geometry. The rectified photos which are not suitable for texturing but necessary
for measuring, can be resized in constrain proportion to real geometry before measuring process.
Height and width of windows, doors, horizontal and vertical distance between windows from upper
left corner of the photo dimensions of doors and windows are parameters that should be measured
to run the program as a plugins in SketchUp Trimble. The system can use these parameters and
texture file names and file paths to create the façade semi-automatically. To avoid leaning geometry
the textures of windows, doors, etc, should be cropped and rectified from perpendicular photos, so
that they can be used in the program to create the whole façade along with its geometries. Texture
enhancement should be done in advance such as removing disturbing objects, exposure setting,
left-right up-down transformation, and so on. In fact, the quality, small data size, scale and semantic
database for each façade are the prominent advantages of this method.
