{"id":1974,"date":"2014-02-05T17:32:44","date_gmt":"2014-02-05T17:32:44","guid":{"rendered":"https:\/\/www.anagram.at\/en\/diplomarbeit\/results\/"},"modified":"2014-02-05T17:32:44","modified_gmt":"2014-02-05T17:32:44","slug":"results","status":"publish","type":"page","link":"https:\/\/www.anagram.at\/en\/diplomarbeit\/results\/","title":{"rendered":"Results"},"content":{"rendered":"


\n
\n Next:<\/b> Calibration<\/a>
\n Up:<\/b> Diplomarbeit<\/a>
\n Previous:<\/b> Summary<\/a>
\n<\/p>\n

<\/p>\n

Results
\n<\/h1>\n<\/p>\n

\n
\nMan darf nicht glauben<\/p>\n

eine Idee k\u00f6nne durch den Beweis ihrer Richtigkeit<\/p>\n

selbst bei gebildeten Geistern Wirkungen erzielen.<\/p>\n

Man wird davon \u00fcberzeugt, wenn man sieht,<\/p>\n

wie wenig Einflu\u00df die klarste Beweisf\u00fchrung<\/p>\n

auf die Mehrzahl der Menschen hat.<\/p>\n

Der unumst\u00f6\u00dfliche Beweis kann von seinem<\/p>\n

gebildeten Zuh\u00f6rer angenommen worden sein,<\/p>\n

aber das Unbewu\u00dfte in ihm wird ihn schnell zu<\/p>\n

seinen urspr\u00fcnglichen Anschauungen zur\u00fcckf\u00fchren.<\/i>
\n[George B. Shaw].<\/p>\n<\/blockquote>\n

\nIn this chapter, the accuracy of the implemented methods, which yield to a 3D reconstruction of the recorded environment, is investigated. First of all the cameras have to be calibrated. Every calibration technique formulates a mathematical model of a camera which is used to solve the equations that relate the known world coordinates and the projected points on the sensor plane. The model of the camera can also be used to re-project the world coordinates to image coordinates. For an estimation of the accuracy, the distance between re-projected coordinates and the coordinates used for calibration is measured.
\nThe next step in the reconstruction is the line detection. If it worked perfectly it would cover all pixels of a straight line with a length more than the minimum length allowed. The results of the iterative line detector using synthetic and real data is presented in Section 4.2<\/a>. Synthetic data has the benefit that an ideal environment and all possible special cases can be modeled. Using real data, the accuracy of the method in a real environment can be investigated.
\nIn Section 4.3<\/a> the correspondence analysis is controlled, also using synthetic and real data. When corresponding lines are found, it can be compared with true values, both in a synthetic and a real world. The synthetic data is modeled using MAYA4.1<\/sup><\/a>. This can be done very easily, because it is possible to let the intrinsic geometry of the real camera and the synthetic camera in maya match. Unfortunately, lens distortion cannot be modeled thus cannot be investigated with synthetic data. The last part of this chapter reviews the 3D reconstruction itself. <\/p>\n<\/p>\n\n

Subsections<\/strong><\/a><\/p>\n