Advanced system model for 1574-nm imaging, scannerless, eye-safe laser radar

Laser radar based on gated viewing uses narrow laser pulses to illuminate a whole scene for direct (incoherent) detection. Due to the time of flight principle and a very fast shutter with precisely controlled delay time, only light reflected in the range R (range slice AR) is detected by a camera. Scattered light which reaches the shutter outside a given exposure time (gate) is suppressed. Hence, it is possible to "look" along the optical axis through changing atmospheric transmissions (rain, haze, fog, snow). For each laser pulse, the grey value image ES(x,y) of the camera is captured by a framegrabber for subsequent evaluation. Image sequences from these laser radar systems are ideally suited to recognize objects, because of the automatic contrast generation of the technology. Difficult object recognition problems, detection, target tracking, or obstacle avoidance at bad weather conditions are favorite applications. In this paper we discuss improvements in the system modelling and simulation of our laser radar system. Formerly the system performance was calculated for the whole system using the signal-to-noise ratio (SNR), leading to a general estimation of the maximum range of target detection. Changing to a pixel oriented approach, we are now able to study the system response for targets with arbitrary two and even three dimensional form. We take into account different kinds of target reflectivity and the Gaussian nature of the illuminating laser spot. Hence it is possible to simulate gray value images (range slices) and calculate range images. This will lead to a modulation transfer function for the system in future. Finally, the theoretical considerations are compared with experimental results from indoor measurements.