Bibliography: pages 167-170. / The purpose of this thesis was to develop a near real time photogrammetric PC based system to study the regional body surface motions of human beings during respiration. By being able to measure the body surface motion of the human torso during breathing it is possible to evaluate the respiratory muscle functions, i.e. the functions of the diaphragm and intercostal muscles, the only skeletal muscles that are essential to life and which enable us to breathe. Previous studies in this field over the past years have successfully employed stereophotogrammetric analysis. The traditional photogrammetric method however was very time consuming in that it involved the taking of stereo pictures with 35mm cameras, developing the negatives and measuring them in a stereocomparator. The time factor involved severely restricted the research to physiologic studies. For the study to be expanded so that it could be used as a diagnostic tool in the hospital ward environment the time to obtain the necessary results had to be considerably reduced and simplified. A near real time photogrammetric PC based system had to be designed to replace the traditional photogrammetric method. A near real time system was developed for capturing dynamic pictures of the breathing cycle of the patients using a pair of video cameras, a two channel split screen vision mixer and a video cassette recorder with digital storage, all mounted on a custom built hospital cot for easy use in the hospital ward. The images to be measured are selected by viewing the split screen in dynamic stereoscopic mode and selected images are captured on an IBM personal computer equipped with an image processing card. A computing package called TAG has been designed for easy use by non-photogrammetric personnel to evaluate stored images. TAG guides the user through the different processing stages from connecting the video machine to the IBM computer, transferring the selected images, measuring control and object points on the images, and processing the data.