To document

Abstract

Enabling automated 3D mapping in forests is an important component of the future development of forest technology, and has been garnering interest in the scientific community, as can be seen from the many recent publications. Accordingly, the authors of the present paper propose the use of a Simultaneous Localisation and Mapping algorithm, called graph-SLAM, to generate local maps of forests. In their study, the 3D data required for the mapping process were collected using a custom-made, mobile platform equipped with a number of sensors, including Velodyne VLP-16 LiDAR, a stereo camera, an IMU, and a GPS. The 3D map was generated solely from laser scans, first by relying on laser odometry and then by improving it with robust graph optimisation after loop closures, which is the core of the graph-SLAM algorithm. The resulting map, in the form of a 3D point cloud, was then evaluated in terms of its accuracy and precision. Specifically, the accuracy of the fitted diameter at breast height (DBH) and the relative distance between the trees were evaluated. The results show that the DBH estimates using the Pratt circle fit method could enable a mean estimation error of approximately 2 cm (7–12%) and an RMSE of 2.38 cm (9%), whereas for tree positioning accuracy, the mean error was 0.0476 m. The authors conclude that robust SLAM algorithms can support the development of forestry by providing cost-effective and acceptable quality methods for forest mapping. Moreover, such maps open up the possibility for precision localisation for forestry vehicles.

Abstract

This paper provides an overview of recent developments in remote and proximal sensing technologies and their basic applicability to various aspects of forest operations. It categorises these applications according to the technologies used and considers their deployment platform in terms of their being space-, airborne or terrestrial. For each combination of technology and application, a brief review of the state-of-the-art has been described from the literature, ranging from the measurement of forests and single trees, the derivation of landscape scale terrain models down to micro-topographic soil disturbance modelling, through infrastructure planning, construction and maintenance, to forest accessibility with ground and cable based harvesting systems. The review then goes on to discuss how these technologies and applications contribute to reducing impacts on forest soils, cultural heritage sites and other areas of special value or interest, after which sensors and methods necessary in autonomous navigation and the use of computer vision on forest machines are discussed. The review concludes that despite the many promising or demonstrated applications of remotely or proximately sensed data in forest operations, almost all are still experimental and have a range of issues that need to be addressed or improved upon before widespread operationalization can take place.

To document

Abstract

Cable yarding is a semi-mechanized timber harvesting system that relies on human machine interaction where a considerable share of the work is done by forest workers. The system is used in mountain forests around the world. Automation of one or more functions could contribute to increased productivity, reduced physical workloads and improved safety conditions. This paper presents a method for sensor data fusion in order to automatically distinguish work phases using Partial Least Squares Discriminant Analysis (PLS-DA). The Robot Operation System (ROS) is implemented to allow for real-time data processing with a maximum latency of 0.16 s. Global Positioning System (GPS), Inertial Measurement Unit (IMU) and camera integration provided a robust solution for 78% correct process segmentation. These results provide a basis for further development from which there is a possibility of expanding this approach for semi-automation, remote control, and autonomous operation.

Abstract

We demonstrate the efficacy of using close-range photogrammetry from a consumer grade camera as a tool in generating high-resolution, three-dimensional coloured point clouds for detailed analysis or monitoring of wheel ruts. Ground-based timber harvesting results in vehicle traffic on 12–70 per cent of the site, depending on the system used, with a variable probability of causing detrimental soil disturbance depending on climatic, hydrological and soil conditions at the time of harvest. Applying the technique described in this article can reduce the workload associated with the conventional manual measurement of wheel ruts, while providing a greatly enhanced source of information that can be used in analysing both physical and biological impact, or stored in a repository for later operation management or monitoring. Approaches for deriving and quantifying properties such as rut depths and soil displacement volumes are also presented. In evaluating the potential for widespread adoption of the method among forest or environmental managers, the study also presents the workflow and provides a comparison of the ease of use and quality of the results obtained from one commercial and two open source image processing software packages. Results from a case study showed no significant difference between packages on point cloud quality in terms of model distortion. Comparison of photogrammetric profiles against profiles measured manually resulted in root mean square errors of between 2.07 and 3.84 cm for five selected road profiles. Maximal wheel rut depth for three different models were 1.15, 0.99 and 1.01 m, and estimated rut volumes were 9.84, 9.10 and 9.09 m3, respectively, for 22.5 m long sections.