Ryohei Masuda

Application of Image Processing Technology for Unloading Automation of Robotic Head-Feeding Combine Harvester*

Abstract The automation of agricultural work is expected to solve several of Japans agriculture problems and help farmers. It has now become necessary to maintain the sustainability of national agricultural production. Thus, we have been trying to develop a robotic combine harvester. While attempts have been made to develop robotic combine harvester and a few prototypes can run and automatically harvest rice and wheat, the unloading task, in which grains are unloaded through the harvesters auger into a grain container on a truck, has not yet been automated. This paper presents a method that uses machine vision to position the harvesters spout at an appropriate point over the grain container on a truck. Experimental results show that this method has sufficient accuracy.

Path-Following Control of a Head-Feeding Combine Robot*

Abstract The study aims to apply an autonomous path-following control for a head-feeding combine robot. A real-time kinematic global positioning system (RTK-GPS) and a GPS compass are used as navigation sensors. These sensors provide the combine robot with the position and heading information required to steer along a target path. To control steering, we applied a Kalman filter that estimates the lateral and heading errors contained in the position, heading, and traveling-speed data. Through field experiments, we demonstrated that the combine robot is capable of automatically following a target spiral path and harvesting rice crops.

Advanced harvesting system by using a combine robot

Abstract The study aims to construct an advanced harvesting system by using a combine robot. The advanced harvesting system means an integral system that monitors and manages the robot remotely. We remodeled a head-feeding combine into a robot so as to harvest rice crops along a target path. A real-time kinematic global positioning system (RTK-GPS) and a GPS compass are used as navigation sensors. These sensors provide the combine robot with the position and heading information required to steer along a target path. To control steering, we applied a Kalman filter that estimates the lateral and heading errors. In addition, a web application was developed to monitor the combine robot remotely while working in the field. We conducted experiments of harvesting rice crop in fields. Through field experiments, we demonstrated that the combine robot is capable of automatically following a target spiral path and harvesting rice crops. At the same time, the information of the position and operational status of the combine was able to be monitored by the web application.

Application of image processing technology for unloading automation of robot combine harvester

Automation of agricultural work has been expected to solve several problems of Japans agriculture and help farmers. Now it is becoming a key to keep sustainability of national agricultural production. Thus we are now trying to develop a robot combine harvester. Attempts have been made to develop robot combine harvester. Although a few prototypes can run and automatically harvest rice and wheat, unloading work, in which grains are unloaded through harvesters auger into a grain container on a truck, has not yet been automated. In this paper, a method using machine vision is presented to position the harvesters spout at an appropriate point. Experimental results show that this method has sufficient accuracy.

Image Processing for Ridge/Furrow Discrimination for Autonomous Agricultural Vehicles Navigation

Abstract Autonomous navigation is important to robotize agricultural vehicles. This paper discusses a development of an image processing method to discriminate between traversable region (furrow) and non-traversable region (ridge) in ridged fields with no crops. Our method consists of 3 parts. First, images were separated into images with shade and images with no shade by gray level histograms. Next, as for images with shade, colour information was used to discriminate between traversable region and non-traversable region. On the other hand, as for images with no shade, texture was focused on. In this research, variance was used as a texture description. The proposed method was tested on images acquired by a camera, taken of different fields and under different light conditions. Our method discriminated between traversable region and non-traversable region successfully.

A Method to Detect the Occurrence of Rice Plant Lodging Using Wavelet Transform

Abstract We propose a method to detect the occurrence of rice plant lodging in paddy field by image processing. Autonomous head-feeding combine needs to have ability to detect the occurrence of rice plant lodging during harvesting. Our method used USB camera image. USB camera is a cheap device and will be easy to be introduced. Wavelet transform was applied to define feature value. Our method could detect the occurrence of rice plant lodging even when light condition changes and when shadow exists in the image.

Turning Performance of Combine Robot by Various Compasses

Abstract The study aims to examine turning performance of a combine robot by various sensors such as GPS compass and geomagnetic compass. A multi global navigation satellite system (GNSS) with a built-in geomagnetic compass is used as navigation sensors. A GPS compass is used for comparison of the geomagnetic compass. These sensors provide the combine robot with the position and heading information required to steer along a target path. To control steering, we applied a Kalman filter that estimates the lateral and heading errors contained in the position, heading, and traveling-speed data while the combine robot is travelling along straight paths. We also applied a switchback turn method when the robot turns at corners. Through field experiments, we demonstrated that the combine robot is capable of automatically following a target spiral path and harvesting rice crops. In addition, the turning performance of the combine robot was evaluated using the GPS compass or the geomagnetic compass. Though the turning performance by the geomagnetic compass is lower than that by the GPS compass, both compasses can provide the combine robot with sufficient accuracy for harvesting rice crops.

Cut-edge and Stubble Detection for Auto-Steering System of Combine Harvester using Machine Vision

Abstract The objective of this study is to detect the cut-edge and the stubble lines of rice plants in the images that are captured with a camera for an auto-steering system of a head-feeding combine harvester. The scene images were acquired with the monochrome CCD camera equipped with an infrared bandpass filter at 850 nm. The camera was mounted on the combines cab. Graphic charts that represent the distribution of the gray level values were used for detection of the cut-edge. However, the cut-edge of rice was not well detected. Next, the stubbles, the short pieces left after the crop is cut, were detected. The averages of the angle error and the offset by stubble line detection were 7.65°, and 11.4 cm, respectively. The system and algorithm can be used for combine steering in rice harvest.