Tofael Ahamed

Development of Auto-Hitching Navigation System for Farm Implements Using Laser Range Finder

Automatic hitching is a challenging task in agricultural automation that could potentially improve safety and reduce drudgery for farmers during coupling of farm implements with tractors. In this study, we attempt the automatic hitching of an implement and tractor using discriminated and localized landmarks and a laser range finder (LRF). Reflectors were used as landmarks, and field experiments were carried out to discriminate among and localize three different shapes of reflector: rectangular, cylindrical, and trapezoidal. The rectangular reflectors were easy to attach to the implement, and accuracy was satisfactory at longer distances. In contrast, the cylindrical and trapezoidal reflectors had the respective drawbacks of fitting data into the regular shape of a circle and template matching of multiple lines from a longer distance. The average errors of rectangular reflectors at the target position were less than 2 cm in x-y coordinates from a distance of 10 m from the LRF. In the second experiment, navigation of the tractor and automatic hitching were done based on localization of reflectors by the LRF, which was investigated using an actual-size autonomous tractor. The results of the experiments confirmed that the autonomous tractor could navigate to the implements position within an average final lateral error of 3 cm and a directional error of 2° for single reflector and double reflector positioning methods on a concrete surface and on a soft, undulating grass field. We found that the number of successful automatic hitching trials was higher with the double-reflector positioning method as compared to the single-reflector positioning method on the concrete surface. For the undulating grass field surface, we attached a guide rail to increase the tolerance of lateral deviation on this surface and noted a higher number of successful trials of automatic hitching with the double reflectors. It can be concluded that automatic hitching is possible using reflectors as an artificial landmark attached to the implement to ensure safe coupling of heavy farm implements with a tractor.

Simple and rapid determination of free fatty acids in brown rice by FTIR spectroscopy in conjunction with a second-derivative treatment.

A simple and rapid method for the determination of free fatty acid (FFA) content in brown rice using Fourier transform infrared spectroscopy (FTIR) in conjunction with second-derivative treatment was proposed. Ground brown rice (10g) was soaked in toluene (20mL) for 30min, and the filtrate of the extract was placed in a 1mm CaF2 liquid cell. The transmittance spectrum of the filtrate was recorded using toluene for the background spectrum. The absorption band due to the CO stretching mode of FFAs was detected at 1710cm(-1), and the Savitzky-Golay second-derivative treatment was performed for band separation. A single linear regression model for FFA was developed using the 1710cm(-1) band in the second-derivative spectra of oleic acid in toluene (0.25-2.50gL(-1)), and the model displayed high prediction accuracy with a determination coefficient of 0.998 and a root mean square error of 0.03gL(-1).

Vision-Based Leader Vehicle Trajectory Tracking for Multiple Agricultural Vehicles.

The aim of this study was to design a navigation system composed of a human-controlled leader vehicle and a follower vehicle. The follower vehicle automatically tracks the leader vehicle. With such a system, a human driver can control two vehicles efficiently in agricultural operations. The tracking system was developed for the leader and the follower vehicle, and control of the follower was performed using a camera vision system. A stable and accurate monocular vision-based sensing system was designed, consisting of a camera and rectangular markers. Noise in the data acquisition was reduced by using the least-squares method. A feedback control algorithm was used to allow the follower vehicle to track the trajectory of the leader vehicle. A proportional–integral–derivative (PID) controller was introduced to maintain the required distance between the leader and the follower vehicle. Field experiments were conducted to evaluate the sensing and tracking performances of the leader-follower system while the leader vehicle was driven at an average speed of 0.3 m/s. In the case of linear trajectory tracking, the RMS errors were 6.5 cm, 8.9 cm and 16.4 cm for straight, turning and zigzag paths, respectively. Again, for parallel trajectory tracking, the root mean square (RMS) errors were found to be 7.1 cm, 14.6 cm and 14.0 cm for straight, turning and zigzag paths, respectively. The navigation performances indicated that the autonomous follower vehicle was able to follow the leader vehicle, and the tracking accuracy was found to be satisfactory. Therefore, the developed leader-follower system can be implemented for the harvesting of grains, using a combine as the leader and an unloader as the autonomous follower vehicle.

Engineering Solutions for Biomass Feedstock Production - Pre-harvest Crop Monitoring System

This work is a part of the integrated research of engineering solution for biomass feedstock production. The goal of this task is to monitor energy crops at the pre-harvest condition using near-real-time remote sensing method. To achieve this elusive goal, the stand alone tower camera platform has been developed to monitor energy crops during plant growing season. A high resolution multispectral camera with a variable lens and pan tilt device are used to capture RGB and CIR images. The system is installed on top of a 38 meter height tower in the field of energy crops. A stand-alone control algorithm has been developed to control the camera gain and exposure time in different natural illumination conditions. The sensing system is remotely controlled using wireless networking. The challenges of this task are the site-specific management; explore the potential of identify the optimum harvesting time (window) and increase the quality assurance of different biomass feedstock. The field layout has been designed to monitor crop growth conditions. The biomass yield and energy content are to be compared using remote sensing models. The correlation with agronomic databases and remote sensing spectral data will be evaluated for the validation and modification of sensing and data processing system.

Preharvest Monitoring of Biomass Production

Preharvest monitoring of biomass production is necessary to develop optimized instrumentation and data processing systems for crop growth, health, and stress monitoring and to develop algorithms for field operation scheduling. Some research questions of specific interest are as follows: (1) What are the major crop sensing needs for energy crop health monitoring and productivity improvement? (2) Which sensor/platform should be used for the field data collection? (3) What is the best process for energy crop data-to-knowledge conversion? In this chapter, we first review the basics of remote sensing and its application to energy crops. We then discuss the development of three near-real-time remote sensing systems, namely, a stand-alone tower-based remote sensing system, close proximity data collection vehicle, and an unmanned aerial vehicle-based remote sensing system to monitor crop growth. The physical status of crop growth and biomass accumulation was projected over the growing seasons. The remote sensing systems included multispectral camera, light detection and ranging (LIDAR), and a global position system sensor. The sensing systems were convenient to perform site-specific monitoring of bioenergy crops and collect data in near real time including ground reference information. These nondestructive measurements included bioenergy crop growth monitoring using typical vegetation index and estimation of biomass yield by correlating it with suitable vegetation index. The field experimental data has been presented to correlate with remote sensing data. To understand the crop growth status over the growing season, the remote sensing data could be correlated with ground truth data to develop a model for predicting dry matter biomass.

On-the-Go Laser-based Sensing System for Measuring Plants Height Using an Autonomous Tractor

Laser-based technologies are recently recognized for their impact on improving agricultural productivity and non-destructive measurement to the environment. The high temporal and low spatial resolutions have the limitations in implementing space and airborne aerial remote sensing methods for site specific management of crops. On the other hand, the ground-based sensing has immense potential for plant growth monitoring and becomes popular for site-specific management. Thus, the objectives of this research were to develop the operational scenario by an autonomous tractor and monitor plant growth based on the reflection height using laser range finders. In this regard, two Laser Range Finders (LRFs) were installed at the back of the autonomous tractor to conduct field experiments. First, LMS 211 LRF for relative positioning to approach the farm implement, coupling the implement and navigated at the field using artificial landmarks. Second, URG 04-LX LRF for quick assessment of crop status. The calibration of laser sensors was conducted for successful landmarks detection and plant canopy sensing. The autonomous tractor was used in this experiment with Programmable Logic Control (PLC) unit, hydraulic actuators for gear shifting and steering mechanism, LRF for relative positioning and RTK GPS for absolute positioning of the tractor. The operational scenario for field operations was developed including coupling, parking, and forward-backward movement using LMS211 LRF. The 2D laser cloud scan points were analyzed to measure reflection height from plants using URG 04-LX LRF. The reflection heights were used as a reference to develop surface map of plant growth under the tree canopy. Different types of plant’s reflection height were measured from the cluster of scan points reflected from the plant canopies at the Agricultural and Forest Research Center, University of Tsukuba. Further experiments will be presented to correlate plant growth and yield of plant biomass over the growing seasons.

Development of Stand Alone Tower Remote Sensing for Energy Crops

Stand alone tower remote sensing platform was developed to monitor energy crops using multispectral imagery. The sensing system was capable of collecting RGB and CIR images of 1920(H)x1080(V) resolution and transferred through wireless networking over the growing season for Miscanthus, Corn, Prairie grass and Switch grass. A motorized variable lens (8-200mm) was implemented to get the high spatial resolution. The system was rotated 0-355o movement in horizontal plane and tilt ±90o movement on vertical plane from a pan-tilt device which was installed with a multispectral camera of 400-1000 nm wavelength on top of a tower at a height of 38 m from the ground. A digital compass was installed with this system to get yaw and pitch of camera position. An algorithm was developed to control automatic image collection in real time for four experimental plots. The calibration for intrinsic parameters of lens distortion was performed for 18, 50, 100, 150 and 200 mm focal length. The maximum spatial resolution using 200 mm focal length at the underneath of the tower with an 89o tilt was 1 mm per pixel. At the corner edge of the field, 270 m away from the tower with an 8o tilt, the spatial resolution was 10 mm per pixel. The minimum spatial resolutions using 18 mm focal length for the corresponding distances and tilt were 15 mm and 10 cm per pixel. In each of the plots, 20 images were collected to develop a site-specific map for energy crops. A 50 mm focal length was selected with 6-40 mm spatial resolutions based on distances and tilt of camera while rotating on horizontal and vertical plane. The temporal resolution was another great advantage which enabled real time image acquisition as opposed to aerial and satellite imagery.

Spectral sensing for dry matter biomass estimation of energy crops

Spectrometry and canopy analyses have been conducted to determine the suitable vegetative indices for estimating the amount of dry matter biomass of energy crops. The ground truth agronomic data were collected to correlate with image information captured from the top of a 38 m tower using multispectral camera on Miscanthus, Corn, Prairie grass and Switch grass. Four 8 m x 8 m plots were selected inside the four hectare field to track the vegetation changes over the growing season. The Vegetation Index (VI), Leaf Area Index (LAI), and intercepted Photosynthetically Active Radiation (PAR) of energy crops have been measured to develop model for predicting dry matter biomass. The central wavelengths for Green (550 nm), Red (670 nm), Red edge (700 and 750 nm), and NIR (800 nm) have been chosen for calculating several indices, NDVI and GNDVI. Color Infrared (CIR) images captured from the tower camera have been analyzed for four experimental plots to correlate ground truth data and the image information. Ortho-rectification and geo-referencing of images have been developed based on the Ground Control Points (GCP) collected by an RTK-GPS unit. The biomass yield and energy content are to be compared using remote sensing models for ground reference data and tower based multispectral images. The correlation with agronomic databases and spectral data is further research to evaluate for the validation and modification of sensing and data processing system.