Modifications of the stereo correspondence problem for passive optoelectronic rangefinder

摘要

The principle of the passive optoelectronic rangefinder (POERF) is known since c. the second half of 1950s. Publications about research and development of POERF do not exist practically. Presumptive causes are presented in this contribution. We deal with POERF research and development since year 2003. The last fully functional demonstration model is from the year 2009. We present firstly its brief description with the aim to explain why modifications of commonly published stereo matching algorithms are necessary. We present a short characteristic of the algorithm that we use at present. In the second part of the contribution, we briefly specify problems pertinent to a suppression of the influence of foreground objects on the accuracy of ranges measurement. The third part indicates the basic approach to the evaluation of POERF accuracy with the aim to define convenient relative characteristics for the precision of estimates of the target range and the corresponding disparity.