SLAM LiDAR Scanner Compared to Mono Camera for Wood Stack Measurement: Accuracy, Efficiency, and Practicality

Introduction

Accurate measurement of wood stacks at roadside storage locations remains a critical operation in forestry. This measurement is vital for timber sales, inventory management, logistical planning, and ensuring fair transactions between forest owners and buyers. Traditional manual methods of wood stack measurement, as outlined in guidelines like the German RVR, are time-consuming, labor-intensive, and prone to human error. This has spurred the forestry sector to explore and adopt advanced optical and laser-based remote sensing technologies to enhance efficiency and precision in wood stack volume determination.

Among these technologies, two prominent approaches have emerged: methods based on mono cameras and those utilizing SLAM (Simultaneous Localization and Mapping) LiDAR scanners. Mono camera systems, often employed in portable devices like smartphones and tablets, offer a cost-effective and accessible solution. These systems typically capture 2.5D measurements by estimating the front surface area of the wood stack from photographs. Conversely, SLAM LiDAR scanners provide true 3D measurements by generating point clouds that capture the entire stack volume.

This article delves into a comprehensive comparison of SLAM LiDAR scanners and mono camera systems for wood stack measurement. Drawing upon a detailed study that evaluated nine different measurement methods, including both mono camera and SLAM LiDAR technologies, we analyze their performance in terms of accuracy, efficiency, and practical applicability. This comparison aims to provide a clear understanding of the strengths and limitations of each technology, helping forestry professionals make informed decisions about adopting the most suitable method for their specific needs. We will explore how these methods compare to traditional manual techniques and discuss the future potential of optical and laser-based solutions in revolutionizing wood stack measurement in the forestry industry.

Understanding Mono Camera and SLAM LiDAR Measurement Methods

To effectively compare SLAM LiDAR scanners and mono cameras, it’s crucial to understand the fundamental principles behind each technology and how they are applied to wood stack measurement.

Mono Camera Systems: 2.5D Measurement Approach

Mono camera systems for wood stack measurement leverage standard RGB cameras, often integrated into smartphones or tablets. These systems operate on a 2.5D measurement principle. They primarily focus on capturing images of the front surface area of the wood stack. The process typically involves:

1. Image Acquisition: The user takes a series of overlapping photographs of the wood stack’s front face, ensuring the device is held parallel to the stack and maintaining a consistent distance. Some systems offer guidance within the app to ensure proper image capture and overlap.
2. Image Stitching: Software within the measurement application stitches these individual photos together to create a comprehensive image of the entire front surface.
3. Contour Detection and Area Calculation: Advanced image processing algorithms, sometimes incorporating machine learning, automatically detect the contour of the wood stack and may even segment individual logs within the image. The front surface area is then calculated.
4. Depth Estimation (Manual Input): Because mono camera systems lack inherent depth perception, the depth of the wood stack (log length) must be manually inputted by the user. This is often achieved by measuring a reference length on the stack or using a pre-determined average log length for the specific timber assortment.
5. Volume Calculation: The gross wood stack volume is estimated by multiplying the calculated front surface area by the manually entered stack depth (log length).

Examples of Mono Camera Systems: Popular mobile applications like Timbeter®, iFOVEA™, and sScale™ utilize mono camera technology for wood stack measurement. These apps offer user-friendly interfaces and automated features for contour detection and volume calculation.

Advantages of Mono Camera Systems:

• Cost-Effective: Utilizes readily available smartphone or tablet cameras, minimizing equipment costs.
• Accessible and Portable: Mobile apps are easy to deploy and use in the field.
• Relatively Simple Operation: The measurement process is generally straightforward and user-friendly.

Limitations of Mono Camera Systems:

• 2.5D Limitation: Only measures the front surface, requiring manual depth input, which can introduce inaccuracies.
• Accuracy Dependence on Reference Scale: Reliant on accurate manual measurement of a reference scale (stack length or reference object), which is susceptible to human error.
• Sensitivity to Stacking Quality and Weather: Measurement accuracy can be affected by uneven stack surfaces, poor stacking quality, and adverse weather conditions (e.g., snow, rain, poor lighting).
• Limited Depth Perception: Cannot directly measure the true 3D volume of the stack, including variations in depth.

Example of a wood stack being measured using the Timbeter app on a smartphone, showcasing the typical workflow of a mono camera-based system.

SLAM LiDAR Scanners: True 3D Measurement Approach

SLAM LiDAR scanners employ Light Detection and Ranging (LiDAR) technology combined with Simultaneous Localization and Mapping (SLAM) algorithms to create detailed 3D representations of the environment. In the context of wood stack measurement, SLAM LiDAR scanners operate as follows:

1. LiDAR Sensor Emits Laser Pulses: The scanner emits rapid pulses of laser light that scan the surrounding environment, including the wood stack.
2. Point Cloud Generation: The scanner measures the time it takes for the laser pulses to return after reflecting off surfaces. This distance information, along with the scanner’s position and orientation, is used to generate a dense point cloud. Each point in the cloud represents a 3D coordinate in space, creating a digital replica of the wood stack and its surroundings.
3. SLAM Algorithm for Positioning and Mapping: The SLAM algorithm simultaneously determines the scanner’s position and orientation in real-time while building the 3D map (point cloud). This is achieved by analyzing the changes in the point cloud as the scanner moves, allowing for accurate and continuous mapping even without GPS in forest environments.
4. 3D Volume Calculation: Software processes the generated point cloud to:
   • Segment the Wood Stack: Isolate the point cloud representing the wood stack from the surrounding terrain and vegetation.
   • Define Volume Boundaries: Establish the top, sides, and, importantly, the bottom boundary of the wood stack for accurate volume calculation. Since the bottom is often obscured, techniques like ground point interpolation may be used to project or estimate the bottom surface.
   • Calculate 3D Volume: Algorithms calculate the volume of the segmented wood stack based on the 3D point cloud data, providing a true 3D gross volume measurement.

Examples of SLAM LiDAR Scanners: The GeoSLAM ZEB HORIZON® is a handheld SLAM LiDAR scanner specifically designed for rapid and accurate 3D mapping in various environments, including forests. Consumer-grade devices like the iPad Pro® with integrated LiDAR sensors also offer SLAM capabilities.

Advantages of SLAM LiDAR Scanners:

• True 3D Measurement: Captures the entire 3D volume of the wood stack, accounting for variations in depth and shape, leading to potentially higher accuracy.
• Depth Perception: Directly measures depth information, eliminating the need for manual depth input and associated errors.
• Less Reliance on Reference Scales: SLAM algorithms inherently provide spatial referencing, reducing the need for external reference scales.
• Potential for Automation: 3D point cloud data allows for more automated segmentation and volume calculation processes.

Limitations of SLAM LiDAR Scanners:

• Higher Equipment Cost: Dedicated SLAM LiDAR scanners or devices like iPad Pro with LiDAR are more expensive than standard smartphones used for mono camera methods.
• More Complex Operation and Data Processing: Scanning process and point cloud data processing can be more complex than mono camera methods, requiring specialized software and expertise.
• Processing Time: Point cloud processing and volume calculation can take longer compared to the near real-time volume estimations of some mono camera apps.
• Line-of-Sight Requirements: LiDAR requires a line of sight to the object being scanned. Dense vegetation or obstructions around the wood stack can hinder data acquisition.
• Bottom Boundary Definition Challenges: Accurately defining the bottom boundary of the wood stack from LiDAR data can be challenging, especially when the bottom is not fully visible.

Example of a wood stack captured by a LiDAR scanner, illustrating the point cloud data and the process of segmenting the stack from its surroundings for volume calculation.

Comparative Performance: Study Results and Analysis

A recent study rigorously compared nine different wood stack measurement methods, including manual, mono camera (iFovea™, Timbeter®, sScale™), and SLAM LiDAR (GeoSLAM ZEB HORIZON®, iPad Pro® with Pix4Dcatch™ and 3D Scanner App™) systems. This study, conducted in a real-world forestry setting, provides valuable insights into the comparative performance of these technologies.

Accuracy and Deviation Analysis

The study measured 47 wood stacks ranging in volume from approximately 9 to 209 cubic meters, totaling around 2700 cubic meters of stacked roundwood. Key findings regarding accuracy and deviation include:

• Significant Variation Across Methods: The study revealed surprisingly significant variations in volume estimations both within and between different measurement methods. Mean relative deviations reached up to 9%.
• Size Dependence of Deviation: The relative deviation was strongly dependent on the size of the wood stack. Smaller stacks exhibited higher relative deviations compared to larger stacks across all methods.
• Underestimation by iPad LiDAR (3D): 3D measurement methods using iPad® RGB and LiDAR (Pix4Dcatch™ and 3D Scanner App™) generally recorded lower timber volumes compared to other methods.
• Overestimation by Handheld LiDAR (SLAM): Conversely, the method based on samples taken with the handheld GeoSLAM ZEB HORIZON® LiDAR tended to overestimate the volume.
• Mono Camera vs. Manual Methods: Mono camera-based methods (iFovea™, Timbeter®, sScale™) and manual section-based methods showed comparable average performance, but with notable variations in individual stack measurements.
• Statistical Significance: Pairwise comparisons using t-tests revealed statistically significant differences between some methods, particularly between iPad 3D Scanner App and Dralle 1 (stereo camera system), indicating systematic differences in their measurements.

Detailed Deviation Ranges:

• Mono Camera Apps (Timbeter, iFovea): Deviations in stack volume estimations using mono RGB-camera methods showed a broad range, potentially due to the reliance on manual reference scale input and sensitivity to operator technique.
• Stereo Camera System (sScale): The stereo camera system (Dralle sScale™) demonstrated relatively consistent performance with a narrower deviation range compared to mono camera apps. The updated version (Dralle 2) showed improved accuracy compared to the older version (Dralle 1).
• SLAM LiDAR (GeoSLAM ZEB HORIZON): While showing a tendency to overestimate volume, the GeoSLAM ZEB HORIZON® exhibited a moderate deviation range. The manual segmentation and processing steps in Cloud Compare software might contribute to some variability.
• iPad LiDAR (Pix4Dcatch, 3D Scanner App): Both iPad LiDAR apps tended to underestimate volume, with a moderate deviation range. The limited range of the iPad LiDAR sensor (5 meters) and potential challenges in capturing the entire stack from all sides might contribute to underestimation.

Pairwise comparisons of different measurement methods against the manual section-based method (SRM), illustrating the relative deviations and the influence of wood stack size on measurement variability.

Efficiency and Practicality Considerations

Beyond accuracy, efficiency and practicality are crucial factors for adopting measurement technologies in forestry operations.

• Time Efficiency: Optical and laser-based methods, including both mono camera and SLAM LiDAR systems, offer significant time savings compared to traditional manual section-based methods. Image acquisition or LiDAR scanning is generally much faster than manual measurements of stack dimensions.
• Operational Complexity: Mono camera apps are generally simpler to operate in the field, requiring minimal training. SLAM LiDAR scanning, while still relatively user-friendly, may require more operator training and attention to scanning technique to ensure complete data capture.
• Data Processing Workflow: Mono camera apps often provide near real-time volume estimations within the app itself, streamlining the workflow. SLAM LiDAR data typically requires post-processing using specialized software (e.g., GeoSLAM HUB, Cloud Compare, Pix4Dcloud, Blender), which can add processing time and require software licenses.
• Equipment Portability and Handling: Mono camera systems using smartphones or tablets are highly portable and easy to handle in forest environments. Handheld SLAM LiDAR scanners like GeoSLAM ZEB HORIZON® are also designed for portability. iPad Pro® with LiDAR, while portable, may be slightly bulkier than smartphones. Stereo camera systems like sScale™ are typically vehicle-mounted, offering efficiency for roadside measurements but less flexibility for accessing stacks in challenging locations.
• Weather and Environmental Conditions:
  • Mono Camera: Performance can be significantly impacted by poor lighting conditions, rain, snow, and fog, which can affect image quality and contour detection.
  • Stereo Camera (sScale): Utilizes artificial LED lighting for nighttime operation, mitigating some lighting limitations. Still susceptible to heavy rain or snow.
  • SLAM LiDAR: Less affected by lighting conditions as LiDAR is an active sensor. GeoSLAM ZEB HORIZON® has an operating temperature range of 0 °C to 50 °C. All optical methods can face challenges in heavy rain or snow. Touchscreen operation in cold and wet conditions can also be a practical limitation.
• Integration with Digital Workflows: Digital measurement methods, including both mono camera and SLAM LiDAR systems, offer the potential for seamless integration into digital forestry workflows. Measurement data can be automatically recorded, geolocated, and transferred to inventory management systems, enhancing data transparency and efficiency in the timber supply chain.

Discussion: Choosing the Right Technology

The study results and comparative analysis highlight that both mono camera and SLAM LiDAR scanners are viable technologies for wood stack measurement, offering improvements over traditional manual methods in terms of efficiency. However, neither technology consistently outperformed the other in terms of accuracy across all conditions. The choice between SLAM LiDAR scanners and mono camera systems depends on a balance of factors including:

• Accuracy Requirements: If the highest possible accuracy is paramount, especially for high-value timber or critical billing processes, SLAM LiDAR scanners offer the potential for more accurate 3D volume measurements. However, the study showed that current implementations of both technologies still exhibit variations, and further refinement of 3D volume calculation methods is needed. For many practical applications, well-executed mono camera measurements can provide acceptable accuracy, particularly for larger stacks.
• Budget and Cost Constraints: Mono camera systems are significantly more cost-effective in terms of equipment. If budget is a primary constraint, mono camera apps provide a readily accessible and affordable entry point into digital wood stack measurement. SLAM LiDAR scanners and devices like iPad Pro with LiDAR represent a higher upfront investment.
• Operational Environment and Conditions: If measurements are frequently conducted in challenging weather conditions (poor lighting, rain, snow), SLAM LiDAR scanners may offer more consistent performance due to their active sensing nature and reduced reliance on ambient light. Mono camera systems are best suited for good lighting conditions and may require additional measures (e.g., lighting equipment) for low-light situations.
• Technical Expertise and Workflow Integration: Mono camera apps are generally easier to deploy and integrate into existing workflows, requiring minimal training. SLAM LiDAR systems may require more technical expertise for scanning, data processing, and integration with data management systems. Organizations with GIS or remote sensing capabilities may find SLAM LiDAR integration more straightforward.
• Stack Size and Quality: For smaller and less uniformly stacked wood stacks, both methods may exhibit higher relative deviations. However, the study suggested that deviations tend to decrease for larger stacks. If primarily measuring large, well-formed stacks, both mono camera and SLAM LiDAR can provide reliable results.

Conclusion and Future Outlook

This comparative analysis demonstrates that both SLAM LiDAR scanners and mono cameras offer valuable tools for modernizing wood stack measurement in forestry. While mono camera systems provide a cost-effective and accessible solution for 2.5D measurement, SLAM LiDAR scanners offer the potential for more accurate true 3D volume determination.

However, the study underscores that no single method currently emerges as definitively superior in all aspects. Both technologies require further development and optimization to enhance accuracy, particularly for 3D volume calculation with LiDAR and reference scale management in mono camera systems. The surprisingly high variability observed across all methods, including manual techniques, highlights the inherent complexities of wood stack measurement and the influence of factors like stacking quality, operator technique, and environmental conditions.

Future research and development should focus on:

• Improving 3D Volume Calculation Algorithms: Refining algorithms for processing LiDAR point clouds to accurately determine wood stack volume, including better handling of bottom boundary definition and airspace estimation.
• Enhancing Automation and Data Processing: Developing more automated workflows for SLAM LiDAR data processing to reduce processing time and technical expertise requirements.
• Addressing Accuracy Limitations in Mono Camera Systems: Exploring techniques to improve the accuracy of reference scale input and mitigate the impact of stacking quality and weather conditions on mono camera measurements.
• Integrating AI and Machine Learning: Leveraging AI and machine learning for improved log segmentation, contour detection, and automated quality assessment in both mono camera and LiDAR-based systems.
• Developing Hybrid Approaches: Investigating hybrid systems that combine the strengths of both mono camera and LiDAR technologies to achieve optimal accuracy, efficiency, and cost-effectiveness.

Ultimately, the forestry sector stands to benefit significantly from the continued advancement and adoption of optical and laser-based wood stack measurement technologies. As these technologies mature and become more refined, they hold the promise of transforming wood measurement from a labor-intensive and error-prone manual process into a more efficient, accurate, and digitally integrated component of the modern forestry supply chain.

References

(The original article’s references should be listed here, formatted in markdown as links or plain text as per the original format.)

(1) Apple Inc. (2021). iPad Pro. Web site.Online | Gscholar
(2) Berendt F, Miguel-Diez F, Wallor E, Blasko L, Cremer T (2021a). Comparison of different approaches to estimate bark volume of industrial wood at disc and log scale. Scientific Reports 11: 15630.CrossRef | Gscholar
(3) Berendt F, Pegel E, Blasko L, Cremer T (2021b). Bark proportion of Scots pine industrial wood. European Journal of Wood and Wood Products 79: 749-752.CrossRef | Gscholar
(4) Berendt F, Wolfgramm F, Cremer T (2021c). Reliability of photo-optical measurements of log stack gross volume. Silva Fennica 55: 1-13.CrossRef | Gscholar
(5) Boberg A, Lilja J (2016). Precision vid travmätning av rundvirke med en fotoinventeringsteknik applicerat i smarta telefoner [Precision of pile-measurement of roundwood with photographic technology applied in smartphones]. Bachelor thesis in Forest Sciences, Swedish University of Agricultural Sciences, Umeå, Sweden. [in Swedish]Gscholar
(6) Christensen R (2021). Innowacyjne metody pomiaru drewna w lesnicwie i przemysle drezwnym [Certified digital photo measurement of raw wood stacks]. In: Proceeding of the Conference “Innovative Timber Measurement Methods for Forestry and Wood Industry”. Poznan University of Life Sciences (Poznan, Poland) 17-19 Nov 2021, pp. 19. [in Polish]Gscholar
(7) Cremer T, Berendt F, Diez F, Wolfgramm F, Blasko L (2021). Accuracy of photo-optical measurement of wood piles. Environmental Sciences Proceedings 3: 90.CrossRef | Gscholar
(8) Cremer T, Blasko L (2017). Analyse der fotooptischen Vermessung von Kiefernstamm- und -industrieholz im Vergleich zum Sektionsraumaß. [Analysis of the fotooptical mensuration of pine stem wood and industrial wood in comparison to manual stack mensuration] Allgemeine Forst und Jagdzeitung 188: 127-139. [in German]CrossRef | Gscholar
(9) De Miguel-Díez F, Tolosana-Esteban E, Purfürst T, Cremer T (2021). Analysis of the influence that parameters crookedness and taper have on stack volume by using a 3D-simulation model of wood stacks. Forests 12: 238.CrossRef | Gscholar
(10) De Miguel-Díez F, Pettenkofer T, Wallor E, Tolosana-Esteban E, Purfürst T, Cremer T (2022). Influence of log and stack parameters on the conversion factors for Norway Spruce wood stacks using a 3D simulation model. International Journal of Forest Engineering 33 (2): 119-128.CrossRef | Gscholar
(11) Dralle (2022). Rechtssicher, präzise und transparent: 3D-Stereo-Technologie für die zeitgemäße Holzwirtschaft [Legally compliant, precise and transparent: 3D stereo technology for the contemporary timber industry]. Dralle A/S, Hörsholm, Denmark, pp. 2. [in German]Online | Gscholar
(12) Fink F (2004). Foto-optische Erfassung der Dimension von Nadelrundholzabschnitten unter Einsatz digitaler, bildverarbeitender Methoden [Photo-optical measurement of softwood log dimension using digital image processing methods]. PhD thesis, Albert-Ludwigs-Universität, Freibug, Germany, pp. 175. [in German]Gscholar
(13) Fonseca MA (2005). The measurement of roundwood: methodologies and conversion ratios. CABI Publishing, Wallingford, UK, Cambridge, USA, pp. 287.Gscholar
(14) FOVEA (2022). FOVEA Quick Start. Treeva GmbH – FOVEA, Uslar, Germany, pp. 14.Online | Gscholar
(15) GeoSLAM (2022). HORIZON product specification. Web site.Online | Gscholar
(16) Gollob C, Ritter T, Kranitzer R, Tockner A, Nothdurft A (2021). Measurement of forest inventory parameters with Apple iPad Pro and integrated LiDAR technology. Remote Sensing 13: 3129.CrossRef | Gscholar
(17) GPL software (2022). CloudCompare version 2.6.1 – user manual. Web site, pp. 181.Online | Gscholar
(18) Gutzeit E, Ohl S, Voskamp J, Kuijper A, Urban B (2011). Automatic wood log segmentation using graph cuts. In: Proceedings of the International Joint Conference “VISIGRAPP 2010 – Computer Vision, Imaging and Computer Graphics: Theory and Applications” (Richard P ed). Angers (France) 17-21 May 2010. Springer, Berlin, Heidelberg, pp. 96-109.Gscholar
(19) Heinzmann B (2017). Studies on the volume determination of industrial plywood of the wood species spruce with the aid of electronic and single-trunk 3D measurement. PhD thesis, University of Sopron, Hungary, pp. 98.CrossRef | Gscholar
(20) Heinzmann B, Barbu MC (2016). Accuracy of photo-optical measurement of industrial timber using the example of FOVEA. Forstarchiv 87: 194-197.Gscholar
(21) Herbon C (2015). Photogrammetric surveying of wood piles on handheld devices. PhD thesis, Otto von Guericke University Library, Magdeburg, Germany, pp. 195.CrossRef | Gscholar
(22) Hollerl H (2021). LogStackPro – Im Vorbeigehen [LogStackPro – In passing]. Forst and Technik 33: 40-44. [in German]Gscholar
(23) Jodlowski K, Moskalik T, Tomusiak R, Sarzynski W (2016). The use of photo-optical systems for measurement of stacked wood. In: Proceedings of the “49th FORMEC Symposium 2016”. Warsaw (Poland) 4-7 Sept 2016. Faculty of Forestry, University of Life Sciences, Warsaw, Poland, pp. 342.Gscholar
(24) Kärhä K, Nurmela S, Karvonen H, Kivinen V-P, Melkas T, Nieminen M (2019). Estimating the accuracy and time consumption of a mobile machine vision application in measuring timber stacks. Computers and Electronics in Agriculture 158: 167-182.CrossRef | Gscholar
(25) Knyaz VA, Maksimov AA (2014). Photogrammetric technique for timber stack volume control. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences 40 (3): 157-162.CrossRef | Gscholar
(26) Kruglov A, Shishko E (2017). Log pile measurement through 3D modeling. In: Proceedings of the “40th International Conference on Telecommunications and Signal Processing (TSP)”. Barcelona (Spain) 5-7 July 2017. IEEE Xplore, pp. 263-266.CrossRef | Gscholar
(27) Laan Labs (2022). 3D Scanner App. Web site.Online | Gscholar
(28) Luetzenburg G, Kroon A, Bjørk AA (2021). Evaluation of the Apple iPhone 12 Pro LiDAR for an application in geosciences. Scientific Reports 11: 22221.CrossRef | Gscholar
(29) Lummitsch S, Findeisen E, Haas M, Carl C (2019). The perspective of optical measurement methods in forestry. In: Proceedings of the Conference “Photonics and Education in Measurement Science 2019”(Rosenberger M, Dittrich P-G, Zagar B eds). Jena (Germany) 17-19 Sept 2019. Jena, Germany 9 (17): 2019-9./19/2019. SPIE, pp. 52.CrossRef | Gscholar
(30) Meagher D (1980). Octree encoding: a new technique for the representation, manipulation and display of arbitrary 3-D objects by computer. Report no. IPL-TR-80-111, Rensselaer Polytechnic Institute, Image Processing Laboratory, Troy, NY, USA, pp. 121.Gscholar
(31) Miguel-Díez F, Purfürst T, Acuna M, Tolosana-Esteban E, Cremer T (2023). Estimation of conversion factors for wood stacks in landings and their influencing parameters: a comprehensive literature review for America and Europe. Silva Fennica 57 (1): 22018.CrossRef | Gscholar
(32) Müller M (2008). Holzaufnahme im BaySF-Logistik-Prozess: sScale von Dralle sorgt für effiziente Kontrollstichproben [sScale from Dralle ensures efficient control sampling – Timber measurement SF logistics process]. LWF aktuell: 26-27. [in German]Gscholar
(33) Neumann LJ (2022). Analyse und Bewertung von Rundholzvermessungsverfahren [Analysis and assessment of roundwood measurement methods]. Master thesis, University of Freiburg, Germany, pp. 99. [in German]Gscholar
(34) Pásztory Z, Heinzmann B, Barbu M-C (2019). Comparison of different stack measuring methods. Siberian Forest Journal 3: 5-13.CrossRef | Gscholar
(35) Pix4D (2022a). PIX4Dcatch: 3D scanner. Web site.Online | Gscholar
(36) Pix4D (2022b). Processing steps. Web site.Online | Gscholar
(37) Praschl C, Zwettler G (2022). Three-step approach for localization, instance segmentation and multi-facet classification of individual logs in wooden piles. In: Proceedings of the “11th International Conference on Pattern Recognition Applications and Methods – ICPRAM 2022”. Online streaming, 3-5 Feb 2022, vol. 1, pp. 683-688.CrossRef | Gscholar
(38) RVR (2021). Rahmenvereinbarung für den Rohholzhandel in Deutschland (RVR) [Framework agreement for the raw timber trade in Germany]. Deutscher Forstwirtschaftsrat e.V., Berlin, Germany, pp. 62. [in German]Online | Gscholar
(39) Spreafico A, Chiabrando F, Teppati Losè L, Giulio Tonolo F (2021). The iPad Pro built-in LIDAR sensor: 3D rapid mapping tests and quality assessment. The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences 43: 63-69.CrossRef | Gscholar
(40) Stängle SM, Dormann CF (2018). Modelling the variation of bark thickness within and between European silver fir (Abies alba Mill.) trees in southwest Germany. Forestry 91: 283-294.CrossRef | Gscholar
(41) Stängle SM, Weiskittel AR, Dormann CF, Brüchert F (2016). Measurement and prediction of bark thickness in Picea abies: assessment of accuracy, precision, and sample size requirements. Canadian Journal of Forest Research 46: 39-47.CrossRef | Gscholar
(42) Stapel B, Ide M (2021). Innowacyjne metody pomiaru drewna w lesnicwie i przemysle drezwnym [Innovative timber measurement methods for the forestry and wood industry]. University of Life Sciences, Poznan, Poland, pp. 85.Gscholar
(43) Ayari T, Hallereau S (2022). Apple iPad Pro Lidar Module. System plus consulting, web site.Online | Gscholar
(44) Tarasena S, Tsahkna A-G (2021). Rozwiazanie zapewniajace przejrzysta i zrównowazona gospodarke lesna [A solution that ensures transparent and sustainable forest management]. In: Proceeding of the Conference “Innovative Timber Measurement Methods For the Forestry and Wood Industry”. Poznan (Poland), 17-19 Nov 2021. [in Polish]Gscholar
(45) Tech FS (2021). Apple LIDAR demystified: SPAD, VCSEL, and Fusion… For Sensing/AIoT Tech and More – Medium, web site.Online | Gscholar
(46) Vogt M, Rips A, Emmelmann C (2021). Comparison of iPad Pro®’s LiDAR and TrueDepth capabilities with an industrial 3D scanning solution. Technologies 9: 25.CrossRef | Gscholar