Search

The paper shows a way how a 3D scanner can be used in a combination with numerical simulations and corresponding analytical results to estimate deformations and stresses in a loaded structural element, such as beams, pipes, shafts, etc. To compare measurement results, a series of numerical simulations were performed for a test object, which was a short pipe that was loaded externally by a force obtained via a manual clamp, which was continuously measured via a force sensor. A general approach to measure strain via a 3D scanner at the object’s surface is explained. However, the surface strain is not sufficient without an analytical consideration to obtain the information what is really occurring inside a loaded structural element, regarding deformations and stresses. Finally, the simulation and the analytical results are compared with the real measurements that were used to calculate strain tensor components, which are used to calculate stress tensor components using the generalized Hooke’s law.

3D scanning is a rapidly growing field, and face scanning is a branch showing a great potential in various applications, but it is still new and little researched. As smartphone availability and functionality grows, the potential for 3D face scanning mobile apps has massively increased as well. Our study is aimed at finding and testing mobile apps that are suitable for 3D face scanning purposes. In our research we scanned two people’s faces using iOS and Android based smartphones. 6 mobile apps were found to be eligible: Bellus 3D, Heges, Scandy Pro, Capture, Trnio, 3D scanner Pro. 3D Scanner Pro were the only Android based app that was found to be suitable for face scanning purposes. Our research shows that based on visual evaluation, the most accurate face 3D models were made by Scandy pro, while Heges was better priced with similar accuracy. In the top three is also Bellus3D, which doesn’t provide the best resemblance to the scanning objects, but it was the easiest to use and process. Our quantitative measurements comparing face proportions show that the best accuracy for male face is obtained by Scandy pro and 3D Scanner Pro, but for female face – Scandy Pro and Capture.

Reverse manufacturing creates another object similar to the existing object. The point cloud data developed with the help of 3D scanning is used for manufacturing of the complex objects. The cost and time of the reverse manufacturing is less than that of the conventional methods. 3D scanner is an important part of 3D printing ecosystem which is playing most important role in latest research in mechanical engineering. The advantage of 3D scanner consists of ease of use, which derives from the shorter scanning duration and the less demanding skill requirement of the operator. Another advantage of the 3D scanning consists in the higher number of acquired surface points which statistically leads to more accurate description of complex parts. For example, mechanical parts like naval vessels, submarines, weapon systems, engines and hulls do not have 3D CAD files and they can be repaired with the help of 3D scanner in a very efficient and easy manner. 3D scanned data is directly used to make changes in programming of artificial intelligent based welding and machining processes. A lot of time and in turn money can be saved using this advanced technique because traditional measurement methods like callipers, rulers etc. consume time and skill. In this research work, reverse manufacturing is applied to propellers of a local aero model manufacturer. It was found that use of 3D scanners in combination with Coordinate Measuring Machine (CMM) helps in measurement of the features with good quality and accuracy in a very short span of time. Results based on comparison of reverse manufacturing with traditional method related to dimensional accuracy and mechanical properties will be presented in full paper.

Domain shift is one of the most salient challenges in medical computer vision. Due to immense variability in scanners’ parameters and imaging protocols, even images obtained from the same person and the same scanner could differ significantly. We address variability in computed tomography (CT) images caused by different convolution kernels used in the reconstruction process, the critical domain shift factor in CT. The choice of a convolution kernel affects pixels’ granularity, image smoothness, and noise level. We analyze a dataset of paired CT images, where smooth and sharp images were reconstructed from the same sinograms with different kernels, thus providing identical anatomy but different style. Though identical predictions are desired, we show that the consistency, measured as the average Dice between predictions on pairs, is just 0.54. We propose Filtered Back-Projection Augmentation (FPBAug), a simple and surprisingly efficient approach to augment CT images in sinogram space emulating reconstruction with different kernels. We apply the proposed method in a zero-shot domain adaptation setup and show that the consistency boosts from 0.54 to 0.92 outperforming other augmentation approaches. Neither specific preparation of source domain data nor target domain data is required, so our publicly released FBPAug (https://github.com/STNLd2/FBPAug) can be used as a plug-and-play module for zero-shot domain adaptation in any CT-based task.

The article discusses the design of an acquisition system for the 3D surface of human arms. The system is composed by a 3D optical scanner implementing stereoscopic depth sensors and by an acquisition software responsible for the processing of the raw data. The 3D data acquired by the scanner is used as starting point for the manufacturing of custom-made 3D printed casts. Specifically, the article discusses the choices made in the development of an improved version of an existing system presented in [1] and presents the results achieved by the devised system.

Most U.S. voters cast hand-marked paper ballots that are counted by optical scanners. Deployed ballot scanners typically utilize simplistic mark-detection methods, based on comparing the measured intensity of target areas to preset thresholds, but this technique is known to sometimes misread “marginal” marks that deviate from ballot instructions. We investigate the feasibility of improving scanner accuracy using supervised learning. We train a convolutional neural network to classify various styles of marks extracted from a large corpus of voted ballots. This approach achieves higher accuracy than a naive intensity threshold while requiring far fewer ballots to undergo manual adjudication. It is robust to imperfect feature extraction, as may be experienced in ballots that lack timing marks, and efficient enough to be performed in real time using contemporary central-count scanner hardware.

This comprehensive survey highlights the state-of-the-art solutions to fingerprint liveness detection across a variety of datasets and scanner models. This chapter includes most algorithms published between 2014 and 2019, which are ranked according to the Average Classification Error (ACE, the average of the statistical Type I and II errors), Error Rate (ER, the ratio of misclassified fingerprints to total fingerprints), or Accuracy Rate (AR, the ratio of correctly classified fingerprints to total fingerprints), for each scanner model in each dataset. Most algorithms surveyed in this chapter were tested on the various LivDet datasets, but other popular datasets such as ATVS and FVC2000 are included in this survey as well. This chapter reviews the LivDet competition series and its progress over time, the various published algorithm performances on all available LivDet datasets (2009–2017), the performance of traditional machine learning algorithms and their variants, and the performance on miscellaneous datasets. This chapter aims to facilitate the research and development of novel liveness classification algorithms through a clear comparison of algorithm performance.

Raw optical coherence tomography (OCT) images typically are of low quality because speckle noise blurs retinal structures, severely compromising visual quality and degrading performances of subsequent image analysis tasks. In this paper, we propose a novel end-to-end cross-domain denoising framework for speckle noise suppression. We utilize high quality ground truth datasets produced by several commercial OCT scanners for training, and apply the trained model to datasets collected by our in-house OCT scanner for denoising. Our model uses the high-resolution network (HRNet) as backbone, which maintains high-resolution representations during the entire learning process to restore high fidelity images. In addition, we develop a hierarchical adversarial learning strategy for domain adaption to align distribution shift among datasets collected by different scanners. Experimental results show that the proposed model outperformed all the competing state-of-the-art methods. As compared to the best of our previous method, the proposed model improved the signal to noise ratio (SNR) metric by a huge margin of \(18.13\,\mathrm dB\) and only required \(25\,\mathrm ms\) for denoising one image in testing phase, achieving the real-time processing capability for the in-house OCT scanner.

This book introduces methods of geoscientific data acquisition using MATLAB in combination with inexpensive data acquisition hardware such as sensors in smartphones, sensors that come with the LEGO MINDSTORMS set, webcams with stereo microphones, and affordable spectral and thermal cameras. The text includes 35 exercises in data acquisition, such as using a smartphone to acquire stereo images of rock specimens from which to calculate point clouds, using visible and near-infrared spectral cameras to classify the minerals in rocks, using thermal cameras to differentiate between different types of surface such as between soil and vegetation, localizing a sound source using travel time differences between pairs of microphones to localize a sound source, quantifying the total harmonic distortion and signal-to-noise ratio of acoustic and elastic signals, acquiring and streaming meteorological data using application programming interfaces, wireless networks, and internet of things platforms, determining the spatial resolution of ultrasonic and optical sensors, and detecting magnetic anomalies using a smartphone magnetometer mounted on a LEGO MINDSTORMS scanner.

Baggage X-ray scanners are one of the most widely used tools for maintaining mass security. Baggage scanners use the same operating principle as their medical counterparts, but the task entrusted to the scanner operator is different from that of the doctor. The scanner operator’s task is to find if there are any dangerous objects in the X-ray image. The operator has to evaluate the shape and kind of material of the scanned objects within a few seconds. Therefore, there is a need for algorithms and methods that analyse such images. This paper presents the dual-energy X-ray scan image segmentation algorithm Adaptive Horizontal Material Classification (AHMC) that classifies materials into multiple classes. The effect is obtained by clusterization of the two-dimensional histograms of low and high energy images using the sliding window method. On the basis of those histograms, local material classes are created. As a result, local classes are combined into global ones, corresponding to the specific material. Our experiments show that the proposed method achieves performance on the same level in comparison to the standard semi-automatic lookup table based methods, but due to its ability to create any number of material classes that every object in an image is made of, outperforms these methods and act as an initial instance segmentation. The generated segmentation is then used in an innovative pseudo-colorization algorithm for X-ray scans.

This work is related to the originally formulated 3D statistical model-based iterative reconstruction algorithm adopted to computed tomography with flying focal spot. This new reconstruction method is based on a continuous-to-continuous data model, where the forward model is formulated as a shift invariant system. The proposed approach resembles the well-known Feldkamp (FDK) algorithm, which cannot be used with a flying focal spot scanner as the paths of the X-rays used are not equi-angularly distributed. In this situation, a so-called “nutating” reconstruction algorithm is usually used, which is based on a rebinning methodology, thus transforming the reconstruction problem to that of a parallel beam system. Our approach has some significant advantages compared with the FBP methods. Moreover, although our method belongs to the category of iterative reconstruction approaches, thanks to the fact that our proposed model is derived as a shift invariant system, it is possible to use an FFT algorithm to accelerate the calculations that have to be performed. Because of this, we can obtain diagnostic images in a time comparable to that of FBP methods. Computer simulations have shown that the reconstruction method presented here outperforms referential FBP methods with regard to the image quality obtained and can be competitive in terms of time of calculation.

Acquiring reliable digital data from the patient is fundamental for accomplishing a correct diagnosis and a trustworthy treatment plan. For that means, clinicians need to obtain 2 types of data: surface scans from patient oral cavity and medical images from the underlying tissue anatomy.

We aim to develop a system to evaluate pouch attachment to and removal from stoma (artificial anus). For preliminary evaluation, we devoted this study to verify if a color/depth camera can accurately capture the stoma shape. Specifically, a 3D scanner was used to take precise images of three stoma models made of human tissue-mimicking gel and with diameters of 3, 6, and 9 cm to obtain the corresponding point clouds. Then, three stoma models were imaged using a color/depth camera to obtain the point clouds. We used the “CloudCompare” software and its implementation of the iterative closest point algorithm to compare the point clouds. Furthermore, the point clouds in the 3D Euclidean space were represented in a polar coordinate system with origin at the center of gravity of the point cloud to obtain the corresponding histograms. Finally, each point cloud and histogram captured by depth camera and 3D scanner were compared. The evaluation results showed that a larger stoma model provides a smaller difference between the point clouds. Nevertheless, the histograms from the three models were similar. Therefore, histogram comparison may be effective for shape recognition of real stoma from point clouds toward the extraction of information for a system to guide pouch management.

Full-guided surgery includes different steps that might lead to complications. The measurement inaccuracy within CBCT scans was evaluated to be up to 0.26 mm. The resolution of the extraoral scanner used to scan the cast models is specified by the manufacturer as less than 20 μm. Deviations in intraoral scans depend on the extension of the optical impression. Matching between the virtual teeth model and the CBCT scans showed an overall deviation of 0.54 mm. Tooth- and gingiva-borne template overall deviations when compared to the planning were found in the vertical deviations up to 2.34 and 2.14 mm. For the entire manufacturing process during full-guided surgery, mean deviations of up to 0.7 mm are possible.

The lack of standardized pipelines for image processing has prevented the application of deep learning (DL) techniques for the segmentation of the aorta in phase-contrast enhanced magnetic resonance angiography (PC-MRA). Furthermore, large, well-curated and annotated datasets, which are needed to create DL-based models able to generalize, are rare. We present the adaptation of the popular nnU-net DL framework to automatically segment the aorta in 4D flow MRI-derived angiograms. The resulting segmentations in a large database (\(> 300\) cases) with normal cases and examples of different pathologies of the aorta provided from a single centre were excellent after post-processing (Dice score of 0.944). Subsequently, we explored the generalisation of the trained network in a small dataset of images (around 20 cases) acquired in a different hospital with another scanner. Without domain adaptation, only with a model trained with the large dataset, the obtained results were substantially worst than with adding a few cases of the small dataset (Dice scores of 0.61 vs 0.86, respectively). The obtained results created good quality segmentations of the aorta in 4D flow MRI, which can later be post-processed to assess blood flow patterns, similarly than with manual annotations. However, advanced domain adaptation schemes are very important in 4D flow MRI due to the large differences in image characteristics between different vendor scanners available in multiple centers.

One of the directions in today’s development of PET scanners is to increase their axial field of view (AFOV). Currently limited to several centimeters, AFOV of the clinically available PET tomographs results in a very low sensitivity (\(\sim \)1%) and requires an extended time for a scan of a whole human body. While these drawbacks are addressed in the so-called, Total Body PET concept (scanner with a significantly elongated field of view), it creates new challenges not only in the mechanical construction but also in the image reconstruction and event selection. The possibility of taking into account of large angle variety of lines of responses (LORs) contributes positively to the sensitivity of the tomograph. However, at the same time, the most oblique LORs have an unfavorable influence on the spatial resolution due to the parallax error and large contribution to the scatter fraction. This forces to determine a new factor - acceptance angle - which is a maximum azimuthal angle for which the LORs are still taken into image reconstruction. Correct determination of such factor is imperative to maximize the performance of a Total Body PET system since it introduces a trade-off between the two main characteristics of scanners: sensitivity and spatial resolution.

Positron Emission Tomography (PET) images are considerably degraded by respiratory and involuntary motions of the patient inside the scanner, having a direct effect in a misdiagnosis. In this paper, a dedicated PET scanner with an open geometry is proposed. This PET configuration poses several challenges to image reconstruction, such as limited angles, motion correction and the sensitivity correction problem. The paper presents a GATE simulation study of image motion correction using XCAT phantom using a multi-frame algorithm called Enhance Multiple Acquisition Frames (EMAF) to correct rigid body and respiratory motion with list-mode data using time of flight (TOF) information and patient motion. This approach is implemented in three phases: frames cutting, image reconstruction and finally image registration. Additionally, the information provided by the TOF is used to improve the reconstruction due to the lack of angular information provided by the proposed open geometry system. Two performance tests are applied to validate the results, obtaining a remarkable resolution improvement after being processed. The peak signal noise ratio (PSNR) values for the corrected and uncorrected images are, respectively, 30 versus 28 dB, and for the image matching precision (IMP), 89% versus 78%. The obtained results show that the method improves the signal intensity over the background in comparison with other literature methods, maximizing the similarity between the ground-truth (static) image and the corrected image and minimizing the intra-frame motion.

Aiming at the problem of on-line inspection and processing due to the alignment errors in the assembly process of large aeronautical components, an adaptive digital processing and assembly method is proposed, and a mobile robot processing system for milling mating surfaces is designed. The preprocessing algorithm of a large number of scattered point cloud data based on the laser profile scanner is studied, and the LSPIA method based on the selection of key points is used to reconstruct the NURBS parametric surface, and the fitting error is analyzed. According to the assembly requirements of multiple sets of mating surfaces at the same time, a calculation method for the minimum machining allowance of each area to be processed is proposed, and the processing track is optimized. Finally, the proposed method is verified by building a prototype of the equipment. The experimental results show that the method meets the technical indicators required by the system, and it can be applied to engineering practice, effectively reducing labor intensity.

The point-to-point distance, the spot size and its shape limit the minimum size of objects that can be spatially resolved in a TLS point cloud. As the laser beam has a footprint of at least a few millimeters, adjacent laser spots overlap if the sampling interval is chosen small. Thus, they do not provide individual information about the object surface and they are correlated. To evaluate the performance of different terrestrial laser scanners to resolve small objects spatially, we investigate their resolution capabilities. Our results show that the expansion and the magnitude of the resolution capability vary between the scanners due to the different focusing and shape of the laser beam, and the rotational speed of the deflecting mirror. Furthermore, we use the resolution capability to assess which scanner provides the most uncorrelated information. Thus, this study provides a measure to judge the scanners’ usability for specific applications, such as finding a crack in a wall.

Due to improved technology terrestrial laser scanners (TLS) are increasingly used for tasks demanding high accuracy, such as deformation monitoring. Within this field, often long-term measurements are acquired, for which the temporal stability of the laser scanner’s observations need to be assured or at least its magnitude and influence factors should be known. While these influence factors have been investigated for most of the geo-sensors taking part in long-term monitoring, it has not been investigated for TLS yet. In this study, we empirically reveal the stability of terrestrial laser scanner observations at long-term measurements. With these investigations, we can analyze the drifts, which occur in the polar observations during the warm-up phase and after it. It is shown that the drifts cause both rigid body movements and inner shape deformations of the point cloud. By re-stationing the scanner during long-term measurements, the drifts in the vertical angle in particular can be reduced by half.

Background The Bayesian penalized likelihood (BPL) algorithm Q.Clear (GE Healthcare) allows fully convergent iterative reconstruction that results in better image quality and quantitative accuracy, while limiting image noise. The present study aimed to optimize BPL reconstruction parameters for 18F-NaF PET/CT images and to determine the feasibility of 18F-NaF PET/CT image acquisition over shorter durations in clinical practice. Methods A custom-designed thoracic spine phantom consisting of several inserts, soft tissue, normal spine, and metastatic bone tumor, was scanned using a Discovery MI PET/CT scanner (GE Healthcare). The phantom allows optional adjustment of activity distribution, tumor size, and attenuation. We reconstructed PET images using OSEM + PSF + TOF (2 iterations, 17 subsets, and a 4-mm Gaussian filter), BPL + TOF (β = 200 to 700), and scan durations of 30–120 s. Signal-to-noise ratios (SNR), contrast, and coefficients of variance (CV) as image quality indicators were calculated, whereas the quantitative measures were recovery coefficients (RC) and RC linearity over a range of activity. We retrospectively analyzed images from five persons without bone metastases (male, n = 1; female, n = 4), then standardized uptake values (SUV), CV, and SNR at the 4th, 5th, and 6th thoracic vertebra were calculated in BPL + TOF (β = 400) images. Results The optimal reconstruction parameter of the BPL was β = 400 when images were acquired at 120 s/bed. At 90 s/bed, the BPL with a β value of 400 yielded 24% and 18% higher SNR and contrast, respectively, than OSEM (2 iterations; 120 s acquisitions). The BPL was superior to OSEM in terms of RC and the RC linearity over a range of activity, regardless of scan duration. The SUVmax were lower in BPL, than in OSEM. The CV and vertebral SNR in BPL were superior to those in OSEM. Conclusions The optimal reconstruction parameters of 18F-NaF PET/CT images acquired over different durations were determined. The BPL can reduce PET acquisition to 90 s/bed in 18F-NaF PET/CT imaging. Our results suggest that BPL (β = 400) on SiPM-based TOF PET/CT scanner maintained high image quality and quantitative accuracy even for shorter acquisition durations.

This paper purpose the design and development of a testing bench to analyze the vibration spectrum of some static elements. In the paper, we are going to describe the application of dynamic vibration analysis into static tests analysis of elements as concrete, glass or wood, the design and methodology of the testing bench and the software development to help the point to point continuous analysis (CP2P) used to evaluate de frequencies in the vibration wave. Finally it is exposed the tests done with those elements, the results and the future research in topics as data size or time to processing. This paper open the vibration analysis research to others application fields different to rotary machinery, in this specific case to a new type of scanner for the products verification.

Background Despite the advent of clinical PET-MR imaging for routine use in 2011 and the development of several methods to address the problem of attenuation correction, some challenges remain. We have identified and investigated several issues that might affect the reliability and accuracy of current attenuation correction methods when these are implemented for clinical and research studies of the brain. These are (1) the accuracy of converting CT Hounsfield units, obtained from an independently acquired CT scan, to 511 keV linear attenuation coefficients; (2) the effect of padding used in the MR head coil; (3) the presence of close-packed hair; (4) the effect of headphones. For each of these, we have examined the effect on reconstructed PET images and evaluated practical mitigating measures. Results Our major findings were (1) for both Siemens and GE PET-MR systems, CT data from either a Siemens or a GE PET-CT scanner may be used, provided the conversion to 511 keV μ-map is performed by the PET-MR vendor’s own method, as implemented on their PET-CT scanner; (2) the effect of the head coil pads is minimal; (3) the effect of dense hair in the field of view is marked (> 10% error in reconstructed PET images); and (4) using headphones and not including them in the attenuation map causes significant errors in reconstructed PET images, but the risk of scanning without them may be acceptable following sound level measurements. Conclusions It is important that the limitations of attenuation correction in PET-MR are considered when designing research and clinical PET-MR protocols in order to enable accurate quantification of brain PET scans. Whilst the effect of pads is not significant, dense hair, the use of headphones and the use of an independently acquired CT-scan can all lead to non-negligible effects on PET quantification. Although seemingly trivial, these effects add complications to setting up protocols for clinical and research PET-MR studies that do not occur with PET-CT. In the absence of more sophisticated PET-MR brain attenuation correction, the effect of all of the issues above can be minimised if the pragmatic approaches presented in this work are followed.

Combining datasets is vital for increased statistical power, especially for neurological conditions where limited data is available. However, variance due to differences in acquisition protocol and hardware limits our ability to combine datasets. We propose an iterative training scheme based on domain adaptation techniques, aiming to create scanner-invariant features while simultaneously maintaining overall performance on the main task. We demonstrate this on age prediction, but expect that our proposed training scheme will be applicable to any feedforward network and classification or regression task. We show that not only can we harmonise three MRI datasets from different studies, but can also successfully adapt the training to work with very biased datasets. The training scheme should, therefore, be applicable to most real-world data scenarios, enabling harmonisation for the task of interest.

The evaluation of white matter lesion progression is an important biomarker in the follow-up of MS patients and plays a crucial role when deciding the course of treatment. Current automated lesion segmentation algorithms are susceptible to variability in image characteristics related to MRI scanner or protocol differences. We propose a model that improves the consistency of MS lesion segmentations in inter-scanner studies. First, we train a CNN base model to approximate the performance of icobrain, an FDA-approved clinically available lesion segmentation software. A discriminator model is then trained to predict if two lesion segmentations are based on scans acquired using the same scanner type or not, achieving a \(78\%\) accuracy in this task. Finally, the base model and the discriminator are trained adversarially on multi-scanner longitudinal data to improve the inter-scanner consistency of the base model. The performance of the models is evaluated on an unseen dataset containing manual delineations. The inter-scanner variability is evaluated on test-retest data, where the adversarial network produces improved results over the base model and the FDA-approved solution.

Standard scanned Color Flow Imaging (CFI) is a common blood flow visualization modality. Despite being introduced more than 30 years ago, this technique is still hampered by the conflicting requirements for either a good image quality or a high frame rate. In fact, good image qualities can only be obtained for frame rates between 10 and 20 Hz, which are unsuitable to show dynamically evolving events. This paper presents a high frame rate imaging modality that, once integrated with CFI, allows to overcome the above limitation. Results characterized by improved quality matched to the capability of properly tracking dynamically evolving flow rates are shown.

Mapping of static objects in an indoor environment is a process to obtain the description of that environment. Simultaneous localization and mapping (SLAM) is an active area of research in mobile robotics and localization technology. To accomplish enhanced performance with low cost sensors, Microsoft Kinect has been mounted on a developed aerial platform in association with Skanect. Skanect transforms the structure sensor Kinect into a less expensive 3D scanner that can create 3D meshes. An experimental result shows how the proposed approach is able to produce reliable 3D reconstruction from the Kinect data.

The primary benefit offered by the particle-bed 3D concrete printing process (3DCP) is the high flexibility in design for manufacture of concrete components with complex geometrical features. Accuracy evaluation has become a critical issue need to be overcome before a more widespread application of the particle-bed 3DCP technique in construction. This study reports an inexpensive image acquisition and processing technique using a flatbed scanner for shape accuracy evaluation of geopolymer specimens made using the particle-bed 3DCP process. A set of image processing algorithms is developed to extract useful shape information from the scanned images without any human intervention. Centroid distance function is used as the shape error representation under the Polar coordinate system for the shape error measurement. A color-labeled map in conjunction with the root mean square error are used for assessment of the shape accuracy. The results show that the developed method can satisfactorily be used for shape accuracy measurement of the particle-bed 3D printed specimen.

Terrestrial laser scanning (TLS) is an important technique to obtain the side view data under the forest canopy and the registration of TLS data captured from different positions is an important step to obtain a complete forest dataset. As commonly used TLS data registration methods are not suitable for the forest scene, this paper presents a registration method of forest TLS data based on orienting and positioning data of smartphone. The coarse transformation parameters between two scans are calculated based on the initial scanner direction and position of each scan, and then used as the input of the Iterative Closest Point (ICP) algorithm for stem position points to get the fine transformation parameters. The experimental results show that this method can achieve accurate alignment of the stem points captured from different sides in case of no artificial targets and poor intervisibility between different scans.