Data Acquisition and 3D Modelling - Speech or Presentation Example

Data Acquisition and 3D Modeling Submitted to, Submitted By, of the Submitted on, [August 5th, DATA ACQUISITION Data acquisition, with respect to 3D modeling, refers to the collection of data that is used to represent a physical object that exists. This data is geometric in nature. Data Acquisition mainly involves three main techniques. These techniques are named as follows: Contact, Non-Contact and Destructive. The resultants of data acquisition process are: “2D cross sectional images and point clouds that define the geometry of an object.” (Pham et al. 2006) METHODS OF DATA ACQUISITION Following is a concise discussion about different methods employed for data acquisition or 3D data acquisition. Laser Ranging Systems These systems acquire 3D data by recording the reflection of projected laser on a surface. This approach is highly useful in reconstruction of long distance depths. Structured Light Methods In this method the 3D data is acquired through the measurement of distortions after projection of light on the object. It is simple to use and is considered better for small distances. Moire Fringe Methods Moire Fringe contour patterns are observed through projected grating onto an object. Accurate depth values can be acquired after the analysis of these patterns. Shape from Shading Methods Analysis of two or more images taken under different lighting conditions with a single camera reveal important information about the orientation of surfaces. This method is not very much suitable for 3D data acquisition but it can provide the basis of hybrid analysis. Passive Stereoscopic Methods The triangulation method is used to infer the surface and their 3D nature in this method. It is known as stereoscopy because it uses two cameras. This is a computationally complex task but now a days it is carried out very successfully with the most modern and sophisticated equipments. This technique is good for apparent physical object reconstruction. Active Stereoscopic Methods The active form of stereoscopic methods use strong source of light to acquire 3D data of a physical object. Laser light source are used typically. It can only be employed for industrial usage with a provision of controlled environment. 3 DIMENSIONAL MODELING 3D or three dimensional modeling refers to the making of a real world object by making use of the actual geographical coordinates of the model. 3D modeling usually refers to the execution of 3D computer graphics by using 3D software. This specialized software functions by creating a set of points in 3Dimensional space. These points are connected together by various geometric shapes. These shapes may merely be a line, a dot, a triangle or a curved surface. There are four important methods upon which 3Dimensional models are created. Polygonal modeling. Three Dimensional models are often created as polygonal models that are textured. In Polygonal Modeling a 3D model is created by connected the points in a 3D space with a line segment. These dots or points are generally referred to as vertices when talked about with reference to space. An avid benefit of this modeling technique is that it can be structured and adapted very conveniently by a computer system. This is owing to the fact that Polygonal models are very flexible. A slight disadvantage that is associated with Polygonal modeling is that it cannot be sued to perfectly structure curves. Primitive modeling. By far, the simplest method of modeling a 3D object is 3D Modeling. This method makes use of pre-defined mathematical/geometric shapes such as cylinder, sphere, cube, cone etc. Using these basic geometrical structures complex three dimensional models are created. Since the basic building blocks of such a modeling technique are pre-defined geometrical structures, their construction is very convenient owing to their forms beings precisely and mathematically defined. The application of the primitive type of 3D Modeling is basically in the 3 Dimensional modeling of technical structures. NURBS modeling. NURBS stands for ‘Non-uniform rational B-spline’ This modeling method is often associated with a popular modeling software named Maya. This modeling technique makes possible the construction of excessively smooth 3D Models. This modeling technique has the ability to create smooth curved surfaces. This way it has an edge over the Primitive modeling that has been discussed above. THE NEED The 3D data acquisition and reconstruction process is of high importance to all those fields in which the reconstructed shape of a physical entity needs to be utilized in any manner. For example a dentist would need 3D data acquisition and reconstruction if he wants to analyze the shape of a specific denture. He may observe the deformations and patterns of decay for his use. APPLICATION AND USAGE There are numerous fields where 3D data acquisition and object reconstruction is employed successfully. Discussion of few is as follows, Medical Industry Based on stereoscopic triangulation methods two of high profile medical equipment companies introduced Handyscan 3D scanners. These scanners are capable of acquiring 3D data of human anatomy. This is done in order to analyze the orthopedic and prosthetic disorders in patients. Despite the complexity of computation these new scanners are capable of performing the required function within 1 minute. 3D Topography Mapping of 3D features of surfaces is known as 3D Topography. 2D satellite images are used to construct 3D surface images of roads and buildings. Elberink (2010) emphasizes and highlights the application of 3D data acquisition for topography. 3D ultrasound computer tomography This is used in the timely detection of breast cancer. The 3D data is acquired through ultra sound projections and then the reconstructed patterns are analyzed for diagnosis. According to Birk et al. (2011), “As todays standard screening methods often fail to diagnose breast cancer before metastases have developed, an earlier breast cancer diagnosis is still a major challenge. To improve this situation, we are currently developing a fully three-dimensional ultrasound computer tomography (3D USCT) system, promising high-quality volume images of the breast.” 3D seismic data acquisition This technology is used by the oil exploration companies to explore under water surfaces for oil reserves. The 3D seismic data acquisition is the latest technology that has overlapped former 2D technology. This 3D technology is much more effective and highlights almost all the surface and depth features. Geographical 3D data acquisition The satellite images and aerial recordings are processed to acquire 3D imagery of the situation at surface. Due to high computational requirements clustered computing is used to compute and reconstruct geometrical objects. Triangulated Irregular Network (TIN): The Triangulated Irregular Network (TIN) model characterizes a surface containing varied slopes, as a set of adjoining, non- intersecting triangles. The set of points through which the triangles are made is known as mass points. The surface inside each triangle is a plane. The careful and accurate selection of mass points results in an accurate surface model. Some of the examples of well-placed mass points are the edges of cliffs, the ground surface or a mountain peak. Grid Model: This is a representation of a surface as a regular raster line based net or grid. This model maintains the surface representation in the form of squares and rectangles. Comparison: The TIN is simple and economical to manage. The data storage efficiency is higher than of the grid. It describes the surfaces at diversified levels of resolution. It is sometimes required that the surface may be examined visually and the network of triangles be controlled manually. With more natural appearance of the terrain under study the grid is an easy to store model. It formulates comparatively smoother models and the integration with raster databases is also easier. Due to its inability to use various grid sizes it does not reflect the varied surface complexities properly. Methods of Selecting Data Points for TIN: Above mentioned methods are basically used to acquire 3D data in the form of points and raster images. There are two basic methods of further picking up the points. The VIP Method: This method gathers points by analyzing the significant breaks on the surface. It works suitably for rough, sharp and pointed surfaces. On curved surfaces the results are not satisfactory. Drop Heuristics: This method analyzes a dense DEM. Afterwards it treats the scenario as an optimization problem and selects the optimized points for the best TIN based representation of the surface. Due to some shortcomings of this technique it is suggested that other photographic images may also be analyzed for points inclusion. Methods of Triangulation for TIN: The selected points act as vertices of triangles to be formulated. There are many ways to construct these triangles through these points. Triangles with 60 degrees are deemed as more suitable. The two main important issues are avoidance of intersection of triangles and correct representation of slopes. Following are some of the methods. 1. Distance Ordering: The distances between all points are first sorted and then the closest pairs are connected. This approach reveals many small triangles. 2. Delaunay Triangulation: This method ensures the selection of all sets of three points that lie on the circumference of individual circles. It is mandatory that the circle contains no points other than the three selected points. The Delaunay triangulation can be achieved through multiple ways. Using Thiessen Polygons. This method produces better 60 degrees triangles. By starting from convex hulls and moving inwards. By simple selection of Delaunay edges and points. Figure 1 poses a sample for better understanding. The Delaunay Vertices The Delaunay Traingulation Figure 1 : The Delaunay Triangulation. (Source: ET Solution Center Index) The main advantages of formulation of Delaunay Triangles are as follows, The precision problems are reduced with Delaunay Triangles. The order of points does not affect the process of triangulation. Ensure the closest distance of points to a node. Delaunay Triangles cannot be merged to form bigger triangles. When divided into smaller triangles they do not produce good shape representations. 3. Using Break lines: This approach exploits the TINs ability to work with the breaks of slope. After the triangulation the ‘breaklines’ are introduced in the network as edges. This approach results in non-Delaunay Triangles. It represents differences of slope in a better way. Figure 2 explains the usage and effect of breaklines.   No break lines Soft break lines Hard break lines The Data The Triangulation The Surface 3D View Figure 2: The usage of breaklines. (Source: ET Solution Center Index) 4. Using Contours The contours are common product of data acquisition techniques. The triangulation is done through the points on contours. The triangles produced this way may be at the same elevation. These are called flat triangles. These triangles are problematic when it comes to surface modeling. The problem of flat triangles can be addressed through, Inclusion of more points to mass points set. Generalization of the contours. Addition of breaklines. Figure 3 below elaborates more. The contours The triangulation with flat triangles The elevation The slope (Flat triangles are indicated as green) Figure 3: The Contour based Triangulation. (Source: ET Solution Center Index) CONCLUSION 3D data acquisition and modeling is a high profile scientific activity that has proven its effectiveness in multiple fields. This area is very much alive and available for future research. The professionals concerned with it don’t see saturation for a long period of time. Currently, the field is being utilized very effectively for the betterment of human kind. Be it the area of medical sciences or the Oil exploration industry, the diversified and versatile 3D data acquisition and modeling techniques have the potential of being used in it with full propensity. The Data Modeling techniques are greatly affected due to the data acquisition methods. The comparison of TIN and Grid (DEN) shows that different data processing mechanisms result in different models. After the comparison of several techniques that are used to apply TIN it can now be said that the results exhibited both, the advantages and disadvantages. Each technique has its significant suitability for respective situations. REFERENCES Marshal D. (1997). Methods of Acquisition. Image Acquisition: 3D Imaging. Retrieved from http://www.cs.cf.ac.uk/Dave/Vision_lecture/node12.html. ‘Features, Software, Implants & Prosthetics’. (2011). 3D Data Acquisition Delivers Orthopedics and Prosthetics Solutions. Medical Designs. Retrieved from http://www.medicaldesignbriefs.com/component/content/article/10225. Birk, M., Hagner, C., Balzer, M., Ruiter, N. V., Hübner, M., Becker, J. (2011). Evaluation of the Reconfiguration of the Data Acquisition System for 3D USCT. International Journal of Reconfigurable Computing. 2011. 9 pages. doi:10.1155/2011/952937. Retrieved from http://www.hindawi.com/journals/ijrc/2011/952937/ Gardner, G. (n.d.). Seismic Imaging.  Society of Exploration Geophysicists. Retrieved from http://sepwww.stanford.edu/sep/biondo/3DSI_frame.html. Elberink, S. O. (2010). Acquisition of 3D Topography. International Institute of Geo-Information Science and Earth Observation. Retrieved from http://www.itc.nl/library/papers_2010/phd/oude_elberink.pdf. Pham, D. T., Eldukhri, E. E., Soroka, A. J. (2006). Intelligent production machines and systems. 2nd I*PROMS Virtual Conference. pp. 254. Retrieved from http://books.google.com/books?id=q6kVM1S3F2AC&dq=data+acquisition+and+3D+modeling&source=gbs_navlinks_s ET Spatial Techniques. (n.d.). Triangulated Irregular Network. ET Solution Center Index. Retrieved from http://www.ian-ko.com/resources/triangulated_irregular_network.htm UC Santa Babara. (2011). The TIN Model. Department of Geology. Retrieved from http://www.geog.ucsb.edu/~good/176b/a14.html Read More