Machine Vision
摘要
Higher organisms make significant use of visual information for safe movement through space. The apparent ease with which this information is safely evaluated and translated into sensorimotor actions is perplexing to some scientists working in the field of image analysis. In vehicle automation, machine vision already plays a dominant role due to the wealth of information contained in image sequences, and this role is growing steadily due to the increasing availability of high-resolution sensors and powerful control units for image evaluation. Last but not least, the fascinating progress in the field of artificial intelligence has caused a technological leap in image analysis in the last decade [38]. The chapter elaborates the image formation process and provides a mathematical formulation thereof. Subsequently, typical operations of image processing are presented. The reconstruction of 3d scene geometry including temporal tracking filters are discussed. Finally, two application examples of machine vision are presented: panoptical segmentation and the use in automatic people movers.