Developing New Tools to Determine Plant Spacing for Precise Agrochemical Application

Advances in the usage of computer imaging, communication technologies and the successful development of new techniques for precision agriculture have facilitated a smart-digital revolution in row crop agriculture in recent years. The use of a yield monitor, variable rate application (VRA) for fertilizer and herbicides, soil property maps and Global Navigation Satellite System (GNSS) technology has enabled the development of computer generated prescription maps for farm management. All these technologies are changing agricultural practices from simple mechanical operations to automated operations implemented by robotic-based systems. The automation of individual crop plant care in vegetable crop fields has increased its practical feasibility and improved efficiency and economic benefit. A systems-based approach is a key feature in the mechanization engineering design via the incorporation of precision sensing techniques. The objective of this study was to design sensing capabilities for implementation to measure plant spacing under different test conditions (California, USA and Andalucía, Spain). Three different optical sensors were used: an optical light curtain transmitter and receiver (880nm), a LiDAR sensor (905 nm), and an RGB camera. An active photoelectric transmission sensor, which contained 3 pairs of optical light curtain transmitters and receivers, evaluated the interruption by the tomato stem of the light curtain between the two devices, and was recorded simultaneously in real-time by a high-speed embedded control system. The LiDAR (model LMS 211 in California and LMS 111 in Spain, from SICK AG) was installed in a vertical orientation in the middle of the mobile platform. Additionally, a RGB spatial mosaicked image was used to adjust the data from the light beam and LiDAR sensor and obtain combined information (RGBD where D is for distance). These sensors were used to properly detect, localize, and discriminate between weed and tomato plants. The use of this detection system may result in a new technique that allows for the automatic control of aggressive weeds and the automation of weeding tools.