Development of a sensing system for automated cotton fruit load and vegetation estimation

Development of a sensing system for automated cotton fruit load and vegetation estimation

Measurements of fruit load and vegetation indices can be useful for monitoring plant growth stages and condition and
informing management decisions (e.g. irrigation volume and timing). However, these measurements are laborious to
collect in broad-acre field situations. A camera-based sensing system has been developed to automatically estimate cotton fruit load and leaf area index in cotton. The system uses three cameras to capture overhead views of the crop canopy and an ultrasonic distance sensor to measure crop height. The captured images are analysed to estimate plant density, flower count and boll count, whilst the height is used to estimate the leaf area index of the crop. Three platforms have been developed to convey the sensing system over the field, two ground-based vehicle configurations (one manual, one motorised), and an overhead system which can be mounted on a centre pivot or lateral move irrigation machine. The groundbased systems were evaluated in the 2010/11, 2011/12 and 2012/13 cotton growing seasons, and the overhead-based sensor in the 2012/13 season. This paper describes the developed system and presents an evaluation of the system in cotton at different crop growth stages and lighting conditions. It is concluded that the plant height can be estimated using non-contact sensors under field conditions and plant density and flower count can be estimated from the top view images. However, the overhead cameras underestimated the boll count as the bolls were generally located lower on the plant.