"Machine readable" crops can improve automatic weeding, say US researchers
Robotic weeding systems could be made more effective if they incorporate one of several vision-based methods to distinguish weed from crop, according to researchers from the universities of California and Arizona.