Use of microwave (radar) to detect soil surface under vegetation

For the height/depth control of the agricultural implements, the soil surface should be measured accurately. A technique using microwave was developed to detect the soil surface through plant and/or vegetation.; The feasibility of using the ground penetrating radar (GPR) was investigated. The distance to target was estimated by measuring the time difference between the transmitted and the received signals. The GPR system was interfaced to a microcomputer via a analog to digital converter. The distance measurement was not affected by the existence of vegetation material in most cases. However, the measurement error was about 2 cm in case of very heavy vegetation of 45 cm in thickness. The soil surface detection algorithm was developed and evaluated in the laboratory and in the field. Although the GPR did not detect the sharp transition, it showed the overall profile of soil surface. In the barley field where the average height of standing barley was 84 cm, the measurement error ranged from 0.5 cm to 6.5 cm.; Another technique using microwave was developed and its performance was evaluated. A dual-frequency continuous-wave radar estimated the distance to target by measuring the phase difference between two different frequencies continuous-waves. Based on the operating principles, a prototype of dual-frequency continuous-wave radar system was designed and constructed with a parabolic cylinder antenna. The dual-frequency continuous-wave radar system output a d.c. voltage according to the phase difference. The bistatic mode operation was adopted since the developed radar system required a high isolation between transmitting and receiving functions. The soil surface detection algorithm was developed as well.; The performance of the radar system was affected by the characteristics of target material such as the moisture content and the surface roughness. The field experiment was conducted in the barley filed where the average height of standing barley was 91.5 cm, the maximum measurement errors were 8.91 cm and 8.44 cm for two different experimental spots.