Title :
Electronic recognition of plant species for machine vision sprayer control systems
Author_Institution :
Fac. of Eng., Regina Univ., Sask., Canada
Abstract :
It is noted that machine vision systems have the potential as sprayer controllers to reduce farm chemical use and to increase the effectiveness of crop spraying operations. The author examines the issues in developing real-time plant recognition algorithms and associated electronic hardware. A very efficient algorithm for distinguishing between broadleaf and grassy plant species is proposed. Preliminary tests on video images of several types of field crops are reported. These tests show the potential for this and related image processing algorithms as plant classifiers in real-time systems
Keywords :
agriculture; computer vision; computerised control; computerised pattern recognition; computerised picture processing; real-time systems; broadleaf series; crop spraying; electronic hardware; field crops; grassy plant species; image processing algorithms; machine vision sprayer control systems; real-time plant recognition algorithms; video images; Automatic control; Chemicals; Control systems; Costs; Crops; Machine vision; Production; Soil; Sorting; Spraying;
Conference_Titel :
WESCANEX '91 'IEEE Western Canada Conference on Computer, Power and Communications Systems in a Rural Environment'
Conference_Location :
Regina, Sask.
Print_ISBN :
0-87942-594-6
DOI :
10.1109/WESCAN.1991.160525
