Hardware-Modellierung und Algorithmus-Optimierung für die dreidimensionale Objekterfassung durch CMOS-Bildsensoren mit Kurzzeitintegration und Laserpulsbeleuchtung
This work analyses a solid state 3D-CMOS camera system for direct time-of-flight image acquisition, which consists of a CMOS imaging sensor, a laser diode module for active narrow infrared laser pulse illumination and all optics for image forming. The investigation is carried out using ideal and real signals. A deterministic model describes the main signal processing. For this the laser pulse function and various integration window function were sampled. The latter were recorded by using the impulse response of a narrow pulse of TFWHM = 63, 44ps. The visual integration windows give a good understanding of differences between ideal and real output function of measurement systems. The influence of noise sources like photon noise, thermal noise and quantization noise was considered. Further investigations related to intensity and time delay of the received laser pulse show how the output voltage noise of the pixel column circuit depends on. The postprocessing algorithm called MDSI was modelled by the sampled laser pulse and integration windows. An extended version of the distance calculation includes an additional time-shift between the laser pulse and the integration window during the first main signal processing. This gives an larger measurement domain. Measurements of the distance-proportional voltage quotient confirm the extended version, which was created in a computer simulation before.