Image Acquiring Method

Equipment

All pictures were taken using a Canon Power Shot A10 digital camera. The resolution is 960x1280 for portrait images and 1280x960 for landscape images. Neither flash nor zoom is used for objects (= touristic sights). Using zoom would mean that no exact distance measure is possible anymore. For details on objects optical zoom is used when necessary, because the distance information is not needed. Positioning is done with GPS if possible. Where no GPS signal with sufficient accuracy is available, a digital kataster is used to determine the position.

General object database

The drawing 1) describes the setup for the image acquiring.

1) image acquiring setup

Z is the central point of the building (the middle of facade). Gamma is of 90°. Alpha and beta are of 30° resp. -30° (if this was not possible due to local conditions, it is denoted in the detailed image description). A1, A2, A3, B1, B2, B3 are possible positions of the person taking the photo. Either the distance A2 - Z or B2 - Z is chosen in the way, that the building fits nicely to the image (this depends on local conditions, such as narrowness of the street and size of the building). For each building either the combination A1, A2, A3, B2 or the combination B1, B2, B3, A2 is chosen (see pictures 2a) and 2b) below). Hence, there exist four viewpoints at most for each object. Pictures were taken under different illumination conditions and at different times of the day. See the list below for details.

2a) different viewpoints, version 1 2b) different viewpoints, version 1
There are at most 24 (4 x 3 x 2) images for each object.

Specialties of TSG-II-5

TSG-II-5 is a subset of TSG-II containing 5 objects, for which we have all possible 24 (combination of 4 viewpoints, 3 different times of the day, 2 different illumination conditions) images.

Positioning Methods

GPS digital kataster TO BE DONE ...