A cherry tomato harvesting robot equipped with a 3-D vision sensor was
constructed for experimental purposes. The 3-D sensor emitted 3 laser
beams with 2 components of wavelengths: a red (wavelength 685 nm) and
an infrared (830 nm). Three position-sensitive devices (PSDs) were us
ed to detect the light beams which were reflected from the crop. The s
hape of the crop was determined by scanning the laser beams. Red ripe
tomatoes were detected based on the ratio of the red to the infrared r
eflected signals. To avoid a collision, the location of the stems and
leaves was also recognized using the image processing software.