Weed Identification using Ultrasonic Sensor in Labview

Abstract

The presence of weeds and pests in the crop field is a common phenomenon. The success of site-specific pest management depends on accurate identification of the pest and weeds in crop field. An innovative low-cost ultrasonic sensor system was developed to detect weeds and bare spots in wild blueberry cropping system. Ultrasonic sensors were mounted besides the rear wheels of the specially designed Farm Motorized Vehicle. Trimble Ag GPS 332 was mounted above the sensors to locate the exact locations of sensor data points for mapping. Custom software interface was developed in Lab View 8.5 to collect and store the sensor data along with DGPS co-ordinates in a laptop computer. The ultrasonic system calibrated using the fixed height objects in the laboratory and vegetation in the wild blueberry fields. Linear regression analysis showed significant relationship between actual heights and sensor heights (R2= 0.98). The survey of the field for weeds and bare spots detection was carried out at a speed of 0.54 m s-1. The height maps were generated in Arc View 3.2 showing weed patches, bare-spots and wild blueberry plants in selected fields.