感性与理性的完美结合：Berthold K.P. Horn教授的寄语

Images can provide an enormous amount of information—in a compact form—w ithout direct connection w ith the things being imaged.We can determ ine properties of ob jects and their shapes from images.Thismakes“photoelectric perception”such an exiting discip line.The f ield includes cases wherewe do not directly obtain an image by manipulating electromagnetic radiation using lensesorm irrors,but by using sensor that provide some function,integralor projection of the shape or texture.By understanding the physics of image formation,or the processofobtaining these projections,we can endeavor to undo the imaging transformation to discern the underlying truth.It is very satisfying,when understanding how themeasurementwe get are related to the physical reality,allows us to recover information about theworld.

Berthold K.P.Horn

麻省理工学院电子工程系教授

美国工程院院士、AAAI会士

Berthold K.P.Horn教授的个人主页：

http://people.csail.m it.edu/bkph/

机器视觉和光电感知方面的部分奠基性经典论文：

http://people.csail.m it.edu/bkph/selected

MIT复制演示实验(1970年)是人工智能领域的一个里程碑：

http://people.csail.m it.edu/bkph/phw_copy_demo.shtm l

开设世界上第一门机器视觉课程，使其成为一个严谨的学科：(www.xing528.com)

http://people.csail.m it.edu/bkph/talkf iles/Closed_Loop_System_Test.pdf

基于自动驾驶汽车的双边巡航控制理论^[1]：

http://people.csail.m it.edu/bkph/Traきc_Flow_Animation

Berthold K.P.Horn教授主讲的MIT机器视觉课程：

http://people.csail.m it.edu/bkph/courses/6801/6801.htm l

Berthold K.P.Horn教授主讲的MIT计算成像课程：

http://people.csail.m it.edu/bkph/courses/6870/6870

【注释】

[1]这张合影拍摄于2018年，被用于博士后课题“无人车双边巡航控制”的新闻报道。