Search

Normalized cross-correlation is the reference approach to carry out template matching on images. When it is computed in Fourier space, it can handle efficiently template translations but it cannot do so with template rotations. Including rotations requires sampling the whole space of rotations, repeating the computation of the correlation each time.This article develops an alternative mathematical theory to handle efficiently, at the same time, rotations and translations. Our proposal has a reduced computational complexity because it does not require to repeatedly sample the space of rotations. To do so, we integrate the information relative to all rotated versions of the template into a unique symmetric tensor template -which is computed only once per template-. Afterward, we demonstrate that the correlation between the image to be processed with the independent tensor components of the tensorial template contains enough information to recover template instance positions and rotations. Our proposed method has the potential to speed up conventional template matching computations by a factor of several magnitude orders for the case of 3D images. [ABSTRACT FROM AUTHOR]

In this paper, we propose a machine learning approach for detecting superficial defects in metal surfaces using point cloud data. We compare the performance of two popular deep learning architectures, multilayer perceptron networks (MLPs) and fully convolutional networks (FCNs), with varying feature sets. Our results show that FCNs (F1=0.94) outperformed MLPs (F1=0.52) in terms of precision, recall, and F1-score. We found that transfer learning with pre-trained models can improve performance when the amount of available data is limited. Our study highlights the importance of considering the amount and quality of training data in developing machine learning models for defect detection in industrial settings with 3D images. [ABSTRACT FROM AUTHOR]

In this paper, an algorithm for macrofractographic study of materials is proposed on the basis of analysis and sampling of functions and means of measurement of a digital optical microscope. It consists of a consistent implementation of methods developed in this work for studying the fracture surface of materials in a 3D image using digital image processing and using automatic lens recognition and proximity functions. To evaluate the applicability of developed algorithms, dispersion-strengthened composites with an aluminum matrix with different particle sizes and volume fractions fractured under tension were studied. On the basis of results of research, features were established in the mechanisms of fracture and locations of foci and crack nucleation zones during tensile tests depending on the amount content of solid phase in the structure of dispersion-hardened composite materials obtained using internal oxidation technology. [ABSTRACT FROM AUTHOR]

An operator feels a burden when he/she controls a rescue robot remotely because he/she has to keep watching camera images to find the target object. We think that this burden can be reduced by the combination of Head Mounted Display (HMD) and object recognition by deep learning. In the first half part of this study, we examine the effect that how presentation method by You Only Look Once (YOLO), a deep learning algorithm, and its recognition results to an operator wearing HMD. In the experiment, three methods of presentation were set: no display of object recognition, display only one object recognition result, and display 80 kinds of object recognition results. Under each presentation method, we measured the time it took for the operator to operate the robot and complete the given task. Additionally, we ask a questionnaire for each experiment. The results of the questionnaire showed that the method to present only one object recognition result was useful. In the second half part of this study, we develop a system to present 3D images with YOLO added, to further ease the burden of object search. Furthermore, we numerically prove that this system represents depth. In the experiment, two methods of displaying were set up: 2D images with Bounding Box (BB) by YOLO and 3D images with BB by YOLO. For each method of presentation, the operator operated the robot and recorded the number of objects found within a time limit. Additionally, we asked a questionnaire at the end of the search in each condition and at the end of all the experiments. The results of the questionnaire suggested points that need to be improved. Furthermore, we consider the flicker of the image found in the experiment. [ABSTRACT FROM AUTHOR]

This paper presents a high-performance 2D-to-3D conversion algorithm technique and its VLSI architecture design for natural scenes. In this study, the depth map was generated based on the horizontal dividing line concept. Also, the dividing line was estimated by obvious color difference on the local region, and the current dividing line was a demarcation from the maximum value to the minimum one for the depth map. A Depth Based Image Rendering technology was employed to generate stereoscopic images according to the image channel of the depth map. Based on the proposed algorithm, a real-time VLSI architecture is presented. Also, for the purpose of cost efficiency, a module-based hardware consisting of a timing schedule control should be designed. The multiplications and divisions of the algorithm can be minimized by circuit design, and so this helps reduce the complexity of this system. Only one frame memory is required to generate real-time 3D images using the depth map. The depth map can be immediately calculated by referring to the location of dividing lines, which can reduce the storage size and I/O bandwidth. The circuit is simulated and verified by one FPGA chip. The critical path is at the multiplex and one flip-flop, and the maximum clock rate can achieve 205 MHz. To be VGA compatible, the stereoscopic RGB pixels can be outputted in parallel per clock to the interface. The overall maximum data rates of the proposed 3D conversion chip can achieve 610 M bytes per second. This can meet the real-time HD requirement. [ABSTRACT FROM AUTHOR]

A method is proposed for the formation and transmission of 3D holographic images, based on forming a 2D holographic signal by subtracting the carrier spatial frequency from it and thereby compressing the holographic information for fast transmission over a standard radio channel in real time. The data on 3D structure of the holographic object is recorded using coherent and incoherent radiation, in which an equivalent of an object hologram is obtained using the structured light techniques, that can be transmitted over a distance as a 2D frame, similar to a hologram with eliminated redundancy. Such transmission of 3D holographic information is similar to the single sideband transmission used in radio engineering, with the subtraction of its carrier frequency and possible subsequent introduction of any carrier at a distance from the holographic object. It can be applied for problems of holographic TV and 3D augmented reality. The article provides examples of this technology used for creating holographic images of a real 3D portrait of a person and transmitting it over a Wi-Fi channel. 3D images formed in the IR range are shown. This technology is easily adaptable to the tasks of integrating images of different parts of the electromagnetic radiation spectrum. Possible applications of the method in 3D instrumentation are proposed. [ABSTRACT FROM AUTHOR]

Recognition of an object from an image or image sequences is an important task in computer vision. It is an important low-level image processing operation and plays a crucial role in many real-world applications. The challenges involved in object recognition are multi-model, multi-pose, complicated background, and depth variations. Recently developed methods have dealt with these challenges and have reported remarkable results for 3D objects. In this paper, a comprehensive overview of recent advances in 3D object recognition using Convolutional Neural Networks (CNN) has been presented. Along with the latest progress in 3D images, general overview of object recognition of 2D, 2.5D, and 3D images is presented. [ABSTRACT FROM AUTHOR]

In this paper we propose a multiscale parametric snake model for ellipse motion estimation across a sequence of images. We use a robust ellipse parameterization based on the geometry of the intersection of a cylinder and a plane. The ellipse parameters are optimized in each frame by searching for local minima of the snake model energy including temporal coherence in the ellipse motion. One advantage of this method is that it just considers the convolution of the image with a Gaussian kernel and its gradient, and no edge detection is required. A detailed study about the numerical evaluation of the snake energy on ellipses is presented. We propose a Newton-Raphson-type algorithm to estimate a local minimum of the energy. We present some experimental results on synthetic data, real video sequences and 3D medical images. [ABSTRACT FROM AUTHOR]

This paper focuses on challenging applications that can be expressed as an iterative pipeline of multiple 3d stencil stages and explores their optimization space on GPUs. For this study, we selected a representative example from the field of digital signal processing, the Anisotropic Nonlinear Diffusion algorithm. An open issue to these applications is to determine the optimal fission/fusion level of the involved stages and whether that combination benefits from data tiling. This implies exploring a large space of all the possible fission/fusion combinations with and without tiling, thus making the process non-trivial. This study provides insights to reduce the optimization tuning space and programming effort of iterative multiple 3d stencils. Our results demonstrate that all combinations that fuse the bottleneck stencil with high halos update cost (>25%, this percentage can be measured or estimated experimentally for each single stencil) and high registers and shared memory accesses must not be considered in the exploration process. The optimal fission/fusion combination is up to 1.65× faster than the case in which we fully decompose our stencil without tiling and 5.3× faster with respect to the fully fused version on the NVIDIA GPUs. [ABSTRACT FROM AUTHOR]

Existing techniques for the simultaneous encryption and compression of three-dimensional (3D) image sequences (e.g., video sequences, medical image sequences) may come with sufficient decryption accuracy or compression ratio, but do not inherently have both; the relationship between them is generally ignored because the images of a sequence are handled individually. To address this problem, we designed Tensor Compressive Sensing (TCS) to simultaneously encrypt and compress a 3D sequence as a tensor rather than several 2D images. To further enhance security, a non-autonomous Lorenz system is constructed to control the three measurement matrices of TCS. The proposed method preserves the intrinsic structure of tensor-based 3D image sequences and achieves a favorable balance of compression ratio, decryption accuracy, and security. Numerical simulation results verify the validity and the reliability of the TCS scheme. [ABSTRACT FROM AUTHOR]

Objective The aim of this study was to assess the role of multislice computed tomography (MSCT) using 2D and 3D images in evaluation of different types of Le Fort fractures. Subjects and methods 30 cases referred from the Emergency unit in Zagazig University Hospital were included in this study, based on clinical evidence of facial trauma, they underwent multislice computed tomography (MSCT) examination and were shown to be positive for pterygoid fractures. All patients subjected to non contrast MSCT in axial cuts. Multiplanar reformatted (MPR) images were acquired using the machine software in sagittal and coronal planes. Finally images were transferred to a workstation for reconstruction of 3D images. Results Sensitivity of multislice computed tomography (MSCT) in detecting compound Le Fort fractures for radiologists A, B and C was 94.4%, 66.7% and 100% respectively. Sensitivity of MSCT in detecting isolated Le Fort fractures for radiologists A, B and C was 100%. The difference between radiologists A, B and C in detecting Le Fort fractures types was insignificant. Good agreement between radiologists A, B, C and intraoperative findings. Conclusion Both 2D-CT and 3D-CT are the best methods in imaging of Le Fort fractures. [ABSTRACT FROM AUTHOR]

We describe a simple technique in which current and freely available technology can be utilised by surgeons while operating the Da Vinci Si/Xi Surgical Robotic systems. This technique allows for a parallel intraoperative display within the surgical console of any desired subject material from a standard computer, utilising commercially available cabling. The ability to view 3D reconstructed images, patient radiology and patient results within the console whilst operating, has the potential to increase operative efficiency, reduce error and aid in adequate resection of tissues. The ease with which our technique is achieved, the benefits of its use and the low cost associated with its implementation support our suggestion that all robotic surgeons incorporate this into their regular operative setup. [ABSTRACT FROM AUTHOR]

To produce realistic synthetic images, it is important to shade objects based on real illumination conditions of a scene. Estimating the direction of the light source of the scene is a key factor to achieve this. Properly estimating this source under noisy conditions is very challenging, and it is a subject of intense research. Computational intelligence techniques offer promising way of tackling this problem. This paper presents a novel neural network-based approach for recovering light source direction in relation to the viewpoint direction of a graphical image in noisy environments. The estimated light source direction can be used for the generation of 3D images from 2D ones. Experiments are performed using both synthetic and real images in noisy scenes. Four synthetic surfaces where generated with varying light source directions for a total of 12 images. Three real images were also used with varying degrees of noise. The experimental results show that the proposed approach is robust and provides a good level of accuracy. [ABSTRACT FROM AUTHOR]

Recent studies have examined various aspects of human physical attractiveness. Attractiveness is considered an evolved psychological mechanism acquired via natural selection because an attractive body reflects an individual's health and fertility. The length of the legs is an often-emphasized aspect of attractiveness and has been investigated using the leg-to-body ratio (LBR), which reflects nutritional status of the infant, health status, fecundity, and other factors that are predictive of physical fitness. However, previous studies of leg length and physical fitness have produced mixed results. The present study investigated the relationship between LBR, defined as the height to perineum divided by total height, and perceived attractiveness. Three-dimensional stimuli (11 male and 11 female) were constructed with various LBR features. Each stimulus was rated by 40 women and 40 men in Japan on a 7-point scale. The results showed that the values closest to the average LBRs were rated as the most attractive. Furthermore, by fitting a quadratic curve on the relationship between attractiveness and LBR, an inverted U-shaped curve with the peak located at the average LBR was observed. In addition, high LBR values were rated as more attractive in females, whereas the opposite was true for males. These results suggest that average LBR is indicative of good health and good reproductive potential, whereas more extreme values are avoided because they could be indicative of diseases and other maladaptive conditions. [ABSTRACT FROM AUTHOR]

The aim of this study was to examine and assess comparative values of HRCT-based multiplanar reformation (MPR), volume rendering (VR) and virtual endoscope built on three-dimensional (3D) shaded-surface display (SSD-based CTVE) for detections of ossicular chain's damage in patients with otitis media. 70 human ears from 70 patients suffering by chronic otitis media or cholesteatoma, who were examined with a preoperative multi-slice computer tomography (MSCT) examination and tympanoplasty in our hospital were collected. The patients ossicular chains were reconstructed with the aforementioned three protocols and assessed via a three-point scoring system by three radiologists. Then, all the patients ossicular chains were reviewed by a surgeon and a radiologist via the same three-point scoring system used during surgeries at same time. By calculation, the Youden's index and coincidence rate were acquired without a significant difference for display of malleus. With regard to the incus, the Youden's index and coincidence rate of VR and MPR did not show any difference, however, both were higher than CTVE. For representation of the stapes, the accuracy of these three modalities is very low; especially, for the CTVE. In conclusion, both MPR and VR are relative robust, and CTVE is not effective for evaluation of small ossicular structures, particularly the stapes. Furthermore, the VR images are real 3D ones. Therefore, it could be the more valuable protocols for detection of the damage of ossicular chain in the patients with otitis media, and should be further applied in the future work. [ABSTRACT FROM AUTHOR]

This paper presents a new framework for an interactive segmentation of 3D images. The framework is based on a bimodal data structure defined by a region adjacency graph (RAG) that is associated with a hierarchical classification tree (HCT). The RAG provides information about the spatial and topological organisation of the extracted regions of the image. The HCT provides information about the similarities between the extracted regions of the image based on a predefined set of features. The first contribution of our work is the combination of a RAG and a HCT. An incremental system was obtained by defining operators that work with and on the RAG and the HCT. If a static predefined processing chain has been defined, these operators can be used in batch mode. If a scheduler is available, they can be used in an adaptive manner. Finally, if a user chooses the operator to be used after each step, the operators can be used interactively. The second contribution of this paper is the formal description of these operators. To give the user the ability to incrementally improve the segmentation, powerful visualisation of the segmentation state and interfaces have been proposed, an important advantage of the proposed framework. To validate the proposed framework, a user study has been conducted in a concrete case of texture segmentation. Our system obtains very satisfactory results even for complex volumetric textures, and helps real users by providing high quality segmentations. The system has been tested by specialists in sonography to segment 3D ultrasound images of the skin. Some examples of segmentation are presented to illustrate the benefit of the interactivity provided by our approach. [ABSTRACT FROM AUTHOR]

Abstract: Objective: To assess the value of multiplanar reformation (MPR) and three-dimensional multidetector computed tomography (MDCT) images in evaluating respiratory distress in neonates. Materials and methods: Sixteen neonates (⩽30days) who presented with respiratory distress (RD) and admitted in the neonatal intensive care unit of the Assuit university children hospital who were selected from a total of 1295 patients presented with RD in whom chest X-ray was inconclusive, so those 16 neonates underwent 64-rows MDCT in the period between November 2010 and November 2011. The recruited patients were sedated before examination and IV contrast medium administration was needed when there was suspected vascular anomaly. Scans were performed in supine position to cover the root of the neck down to the level of the adrenal glands. The images were sent to the workstation to be reviewed in the multiplanar and 3D images. Results: The final radiologic diagnoses included cystic adenomatoid malformation (n= 4, 25%), congenital lobar emphysema (n =1, 7%), pneumonia (n=5, 36%), hypoplastic lung (n =3, 22%), diaphragmatic hernia (n =2, 14%) and postoperative hydro-pneumothorax (n =1, 7%). Conclusion: The combination of fast speed, high spatial resolution, and enhanced quality of MPR and 3D images makes MDCT an ideal non-invasive method for evaluating surgical causes of neonatal respiratory distress. [Copyright &y& Elsevier]

Open cell foams are a class of modern materials which is interesting for a wide variety of applications and which is not accessible to classical materialography based on 2d images. 3d imaging by micro computed tomography is a practicable alternative. Analysis of the resulting volume images is either based on a simple binarisation of the image or on so-called cell reconstruction by image processing. The first approach allows to estimate mean characteristics like the mean cell volume using the typical cell of a random spatial tessellation as model for the cell shape. The cell reconstruction allows estimation of empirical distributions of cell characteristics. This paper summarises the theoretical background for the first method, in particular estimation of the intrinsic volumes and their densities from discretized data and models for random spatial tessellations. The accuracy of the estimation method is assessed using the dilated edge systems of simulated random spatial tessellations. [ABSTRACT FROM AUTHOR]

Fractures of the proximal humerus are typical osteoporotic fractures of the elderly with an increasing incidence. Computed tomography (CT) with 3D reconstruction plays a more and more decisive role in the diagnostics because of an improved understanding of fractures in 3D images. The resulting correct fracture classification has significance for the decision of the best therapy procedure. Currently an extended version of the Codman classification with its four and more fragments is used to give additive information about varus or valgus dislocation, impression or distraction. The comparison of conservative and operative treatment showed no predominance of one of the procedures, therefore both strategies are justified. An operative treatment is recommended in complex fractures. Thus, early functional mobilization, early self-dependence and return to the activities of daily living are possible. Intramedullary nailing is advised in 2-part fractures, in dislocated multipart fractures locking plating or a primary reverse shoulder arthroplasty with refixation of the tuberosities in patients over 75 years. However, the improved plate and nail systems with polyaxiality, calcar screws, modern plate and nails designs as well as the possibility of arthroscopy-assisted nail implantation or plate removal combined with arthroscopic arthrolysis are innovative. [ABSTRACT FROM AUTHOR]

Quantitative evaluation of axial vertebral rotation is essential for the determination of reference values in normal and pathological conditions and for understanding the mechanisms of the progression of spinal deformities. However, routine quantitative evaluation of axial vertebral rotation is difficult and error-prone due to the limitations of the observer, characteristics of the observed vertebral anatomy and specific imaging properties. The scope of this paper is to review the existing methods for quantitative evaluation of axial vertebral rotation from medical images along with all relevant publications, which may provide a valuable resource for studying the existing methods or developing new methods and evaluation strategies. The reviewed methods are divided into the methods for evaluation of axial vertebral rotation in 2D images and the methods for evaluation of axial vertebral rotation in 3D images. Key evaluation issues and future considerations, supported by the results of the overview, are also discussed. [ABSTRACT FROM AUTHOR]

The aim of this paper is to provide a complete overview of the existing methods for quantitative evaluation of spinal curvature from medical images, and to summarize the relevant publications, which may not only assist in the introduction of other researchers to the field, but also be a valuable resource for studying the existing methods or developing new methods and evaluation strategies. Key evaluation issues and future considerations, supported by the results of the overview, are also discussed. [ABSTRACT FROM AUTHOR]

Mining discriminative spatial patterns in image data is an emerging subject of interest in medical imaging, meteorology, engineering, biology, and other fields. In this paper, we propose a novel approach for detecting spatial regions that are highly discriminative among different classes of three dimensional (3D) image data. The main idea of our approach is to treat the initial 3D image as a hyper-rectangle and search for discriminative regions by adaptively partitioning the space into progressively smaller hyper-rectangles (sub-regions). We use statistical information about each hyper-rectangle to guide the selectivity of the partitioning. A hyper-rectangle is partitioned only if its attribute cannot adequately discriminate among the distinct labeled classes, and it is sufficiently large for further splitting. To evaluate the discriminative power of the attributes corresponding to the detected regions, we performed classification experiments on artificial and real datasets. Our results show that the proposed method outperforms major competitors, achieving 30% and 15% better classification accuracy on synthetic and real data respectively while reducing by two orders of magnitude the number of statistical tests required by voxel-based approaches. [ABSTRACT FROM AUTHOR]

Combining helical volumetric CT acquisition and thin-slice thickness during breath hold provides an accurate assessment of both focal and diffuse airway diseases. With multiple detector rows, compared with single-slice helical CT, multislice CT can cover a greater volume, during a simple breath hold, and with better longitudinal and in-plane spatial resolution and improved temporal resolution. The result in data set allows the generation of superior multiplanar and 3D images of the airways, including those obtained from techniques developed specifically for airway imaging, such as virtual bronchography and virtual bronchoscopy. Complementary CT evaluation at suspended or continuous full expiration is mandatory to detect air trapping that is a key finding for depicting an obstruction on the small airways. Indications for CT evaluation of the airways include: (a) detection of endobronchial lesions in patients with an unexplained hemoptysis; (b) evaluation of extent of tracheobronchial stenosis for planning treatment and follow-up; (c) detection of congenital airway anomalies revealed by hemoptysis or recurrent infection; (d) detection of postinfectious or postoperative airway fistula or dehiscence; and (e) diagnosis and assessment of extent of bronchiectasis and small airway disease. Improvement in image analysis technique and the use of spirometrically control of lung volume acquisition have made possible accurate and reproducible quantitative assessment of airway wall and lumen areas and lung density. This contributes to better insights in physiopathology of obstructive lung disease, particularly in chronic obstructive pulmonary disease and asthma. [ABSTRACT FROM AUTHOR]

Background: The volumetric ratio of the tongue to the oral cavity has been recognized to be one of the important factors for the maintenance of stable occlusion. Oral cavity capacity is changed after orthognathic surgery in patients with mandibular prognathism; however, the volumetric changes of the oral cavity including the tongue before and after surgery have not been analyzed. The purpose of this study was to evaluate the morphological and volumetric changes of the tongue and oral cavity following orthognathic surgery using a newly developed vinyl polysiloxane impression method. Materials and methods: The study was performed in fifteen subjects who underwent surgical orthognathic treatment. Impressions of the tongue together with the oral cavity were obtained before orthognathic surgery and 1, 3, and 6 months after orthognathic surgery. These impression patterns were scanned using cone-beam computed tomography (CT), and three-dimensional (3D) images of the oral cavity including the tongue, and the upper and lower dental arches were reconstructed. The morphological and volumetric changes in the oral cavity capacity and the tongue volume were examined. Results: The volume of the tongue with the volume of the oral cavity decreased after orthognathic surgery. There was a correlation between the decrease in the oral cavity capacity and tongue volume. The volumetric ratio of the tongue to the oral cavity seems to be maintained before and after orthognathic surgery. Conclusion: VPS method, free from radiation exposure, may be useful for investigating the morphological and volumetric changes of the tongue and oral cavity, which may possibly influence the stability of the dental arch and occlusion during surgical orthodontic treatment. [ABSTRACT FROM AUTHOR]