Image Processing & Applications

Minimization techniques are widely used for retrieving a 3D surface starting from a single shaded image i.e., for solving the shape from shading problem. Such techniques are based on the assumption that expected surface to be retrieved coincides with the one that minimize a properly developed functional, consisting of several contributions. Among the possible contributes defining the functional, the so called "smoothness constraint" is always used since it guides the convergence of the minimization process towards a more accurate solution. Unfortunately, in areas where actually brightness changes rapidly, it also introduces an undesired over-smoothing effect. The present work proposes two simple yet effective strategies for avoiding the typical over-smoothing effect, with regards to the image regions in which this effect is particularly undesired (e.g., areas where surface details are to be preserved in the reconstruction). Tested against a set of case studies the strategies prove to outperform traditional SFS-based methods.

Keywords: 3D model , Minimization , Shape from shading , Smoothing , Smoothness constraint , Variational approach

This paper presents a comparison of color spaces for material classification. The study includes three device-independent (CIELAB, CIELUV, and CIE XYZ) and seven device-dependent spaces (RGB, HSV, YIQ, YUV, YCbCr, Ohta's I1I2I3, and RG-YeB-WhBl). The pros and cons of the different spaces and the procedures for converting color data among them are discussed in detail. An experiment based on 12 different image data sets was carried out to comparatively evaluate the performance of each space for material classification purposes. The results showed that CIELAB markedly outperformed the other spaces followed by HSV and CIELUV. Conversely, CIE XYZ came out as the worst performing space. Interestingly, no significant difference emerged among the performance of the other device-dependent spaces.

Keywords: classification , color spaces , materials , visual appearance

The subject of this study is the use of local multi-dimensional patterns for image classification. The contribution is both theoretical and experimental: on the one hand the paper introduces a complete and general mathematical model for encoding multi-resolution, rotation-invariant local patterns; on the other experimentally evaluates the use of multi-resolution patterns for image classification both from an information- and performance-based standpoint. The results indicate that the joint multi-resolution model proposed in the paper can actually convey an additional amount of information with respect to the marginal model; but also that the marginal model (i.e. concatenation of features computed at different resolutions) can be a good enough approximation for practical applications.

Keywords: Image classification , Local binary patterns , Rotation-invariant multi-resolution patterns , Texture spectrum

Automatic recognition of different tissue types in histological images is an essential part in the digital pathology toolbox. Texture analysis is commonly used to address this problem; mainly in the context of estimating the tumour/stroma ratio on histological samples. However, although histological images typically contain more than two tissue types, only few studies have addressed the multi-class problem. For colorectal cancer, one of the most prevalent tumour types, there are in fact no published results on multiclass texture separation. In this paper we present a new dataset of 5,000 histological images of human colorectal cancer including eight different types of tissue. We used this set to assess the classification performance of a wide range of texture descriptors and classifiers. As a result, we found an optimal classification strategy that markedly outperformed traditional methods, improving the state of the art for tumour-stroma separation from 96.9% to 98.6% accuracy and setting a new standard for multiclass tissue separation (87.4% accuracy for eight classes). We make our dataset of histological images publicly available under a Creative Commons license and encourage other researchers to use it as a benchmark for their studies.

A family of 26 non-parametric texture descriptors based on Histograms of Equivalent Patterns (HEP) has been tested, many of them for the first time in remote sensing applications, to improve urban classification through object-based image analysis of GeoEye-1 imagery. These HEP descriptors have been compared to the widely known texture measures derived from the gray-level co-occurrence matrix (GLCM). All the five finally selected HEP descriptors (Local Binary Patterns, Improved Local Binary Patterns, Binary Gradient Contours and two different combinations of Completed Local Binary Patterns) performed faster in terms of execution time and yielded significantly better accuracy figures than GLCM features. Moreover, the HEP texture descriptors provided additional information to the basic spectral features from the GeoEye-1's bands (R, G, B, NIR, PAN) significantly improving overall accuracy values by around 3%. Conversely, and in statistic terms, strategies involving GLCM texture derivatives did not improve the classification accuracy achieved from only the spectral information. Lastly, both approaches (HEP and GLCM) showed similar behavior with regard to the training set size applied.

Keywords: GeoEye-1 , Histograms of equivalent patterns , OBIA , Texture

The size distribution of aggregates has direct and important effects on fundamental properties of construction materials such as workability, strength and durability. The size distribution of aggregates from construction and demolition waste (C&D) is one of the parameters which determine the degree of recyclability and therefore the quality of such materials. Unfortunately, standard methods like sieving or laser diffraction can be either very time consuming (sieving) or possible only in laboratory conditions (laser diffraction). As an alternative we propose and evaluate the use of image analysis to estimate the size distribution of aggregates from C&D in a fast yet accurate manner. The effectiveness of the procedure was tested on aggregates generated by an existing C&D mechanical treatment plant. Experimental comparison with manual sieving showed agreement in the range 81-85%. The proposed technique demonstrated potential for being used on on-line systems within mechanical treatment plants of C&D.

Keywords: Construction and demolition waste , Image analysis , Particle size distribution , Texture

This paper presents the advancements of an automatic segmentation procedure based on the concept of Hierarchical Space Partitioning. It is aimed at tolerance inspection of electromechanical parts produced by injection moulding and acquired by laser scanning. After a general overview of the procedure, its application for recognising cylindrical surfaces is presented and discussed through a specific industrial test case.

Keywords: Computer aided tolerancing & inspection , Hierarchical space partitioning , Injection moulding , Reverse engineering

Three-dimensional deformation analysis of human organs is very important from both diagnostic and therapeutic point of view. For example, comparing the deformation field in healthy and pathologic cardiac walls in the systolic phase allows to gather early and accurate information on the onset of heart diseases. MRI tagging is utilized in medicine to visualize with a great deal of detail the structure and morphology of tissues. The tagging process introduces a volumetric system of planes of reference similar to the process of introducing a grating in the 3-D moiré method in transparent media. The paper will analyze the kinematics of 3-D deformation fields and the fundamental concepts involved in 3-D deformation analysis within the restrictions imposed by the MRI method thus providing solutions for the inherent shortcomings encountered in the MRI tagging technique.

Keywords: Deformation of cardiac wall , Digital moiré , Large deformation analysis , MRI tagging , Rotations

Rolling contact wear/fatigue tests on wheel/rail specimens are important to produce wheels and rails of new materials for improved lifetime and performance, which are able to operate in harsh environments and at high rolling speeds. This paper presents a novel non-invasive, all-optical system, based on a high-speed video camera and multiple laser illumination sources, which is able to continuously monitor the dynamics of the specimens used to test wheel and rail materials, in a laboratory test bench. 3D macro-topography and angular position of the specimen are simultaneously performed, together with the acquisition of surface micro-topography, at speeds up to 500 rpm, making use of a fast camera and image processing algorithms. Synthetic indexes for surface micro-topography classification are defined, the 3D macro-topography is measured with a standard uncertainty down to 0.019 mm, and the angular position is measured on a purposely developed analog encoder with a standard uncertainty of 2.9°. The very small camera exposure time enables to obtain blur-free images with excellent definition. The system will be described with the aid of end-cycle specimens, as well as of in-test specimens.

A fast and reliable optical setup is here presented for in-focus imaging of objects in the macro range. The setup uses a camera equipped with an objective embedding a liquid lens, whose focal length is voltage-controlled. The defocus condition of the image is controlled by means of two indexes, both suitable for coarse and for fine adjustments. A purposely designed algorithm makes use of the two indexes, switching from one to the other to position the image in focus by adequately controlling the liquid lens focal length. The setup has been calibrated by means of target planes of known contrasts, and applied to process biomedical images such as fingerprints.

Innovation of fashion-related products implies the continuous search for new and appealing shapes and materials in a short period of time due to the seasonality of the market. The design and manufacturing of such products have to deal with a dimensional variability as a consequence of the new shapes. An additional difficulty concerns properly forecasting the technological behaviour of the new materials in relation to the manufacturing process phases. The control of dimensional variations requires time and resource intensive activities. Human's manual and visual inspection solutions are more common than automatic ones for performing such control, where skilled operators are typically the only ones capable of immediately facing non-standard situations. The full control of such variations is even more subtle and mandatory in the field of spectacles, which are fashion-related products and also medical devices. This paper describes an inspection system developed to monitor the dimensional variations of a spectacles frame during the manufacturing process. We discuss the methodological approach followed to develop the system, and the experimental campaign carried out to test its effectiveness. The system intends to be an alternative to current inspection practices used in the field, and also to provide a methodological approach to enable engineers to systematically study the correlations existing among the frame main functional and dimensional parameters, the material behaviour and the technological variables of the manufacturing process. Hence, the system can be considered a method to systematically acquire and formalise new knowledge. The inspection system consists of a workbench equipped with four high-quality commercial webcams that are used to acquire orthogonal-view images of the front of the frame. A software module controls the system and allows the automatic processing of the images acquired, in order to extract the dimensional data of the frame which are relevant for the analysis. A case study is discussed to demonstrate the system performances.

Keywords: Eyewear industry , Image processing , Inspection systems , Knowledge-based engineering , Product variability

The paper describes a system to supply meaningful insights to designers during the concept generation of new car interiors. The aim of the system is to capture the movements of car passenger's and making these acquisitions directly available as a generative input. The system has been developed by integrating the Abstract Prototyping technique with Motion Capture technology. In addition, a systematic procedure allows the treatment of the collected data to obtain a graphical representation, which can be easily used with standard NURBS-based modeling software. The effectiveness of the system has been evaluated through a testing session conducted with subjects. The outcomes of the testing sessions have highlighted benefits and limitations of the implemented system.

Keywords: Abstract prototyping , human factors , motion capture

Color matching of fabric blends is a key issue for the textile industry, mainly due to the rising need to create high-quality products for the fashion market. The process of mixing together differently colored fibers to match a desired color is usually performed by using some historical recipes, skillfully managed by company colorists. More often than desired, the first attempt in creating a blend is not satisfactory, thus requiring the experts to spend efforts in changing the recipe with a trial-and-error process. To confront this issue, a number of computer-based methods have been proposed in the last decades, roughly classified into theoretical and artificial neural network (ANN)-based approaches. Inspired by the above literature, the present paper provides a method for accurate estimation of spectrophotometric response of a textile blend composed of differently colored fibers made of different materials. In particular, the performance of the Kubelka-Munk (K-M) theory is enhanced by introducing an artificial intelligence approach to determine a more consistent value of the nonlinear function relationship between the blend and its components. Therefore, a hybrid K-M+ANN-based method capable of modeling the color mixing mechanism is devised to predict the reflectance values of a blend.

Keywords: computer-based color assessment , fabric blend , Kubelka-Munk theory , neural networks , spectrophotometer

The continuous progress of interaction technologies reveals that we are witnessing a revolution that is leading to a redefinition of the concept of “user interface” and to the development of new ways to interact with the electronic devices of all sizes and capabilities. Current trends in research related to the Human-Machine Interaction (HMI) show a considerable interest toward gesture, motion-based and full-body based interactions. In this context, a User-Centered Design (UCD) methodology to implement these novel interaction paradigms into consumer products is proposed with the aim to improve its usability, intuitiveness and experience. A case study is used to validate the methodology and measure the achieved improvements in user performance.

Keywords: Design methods , Gesture interaction , User interfaces , User-Centered design

This paper presents the implementation of a new boarding strategy that exploits passenger and hand-luggage detection and classification to reduce the boarding time onto an airplane. A vision system has the main purpose of providing passengers data, in terms of agility coefficient and hand-luggage size to a seat assignment algorithm. The software is able to dynamically generate the passenger seat that reduces the overall boarding time while taking into account the current airplane boarding state. The motivation behind this work is to speed up of the passenger boarding using the proposed online procedure of seat assignment based on passenger and luggage classification. This method results in an enhancement of the boarding phase, in terms of both time and passenger experience. The main goal of this work is to demonstrate the usability of the proposed system in real conditions proving its performances in terms of reliability. Using a simple hardware and software setup, we performed several experiments recreating a gate entrance mock up and comparing the measurements with ground truth data to assess the reliability of the system.

Keywords: Agility measurement , Boarding strategy , Computer vision

Ultrasound is by far the most adopted method for safe screening and diagnosis in the prenatal phase, thanks to its non-harmful nature with respect to radiation-based imaging techniques. The main drawback of ultrasound imaging is its sensitivity to scattering noise, which makes automatic tissues segmentation a tricky task, limiting the possible range of applications. An algorithm for automatically extracting the facial surface is presented here. The method provides a comprehensive segmentation process and does not require any human intervention or training procedures, leading from the output of the scanner directly to the 3D mesh describing the face. The proposed segmentation technique is based on a two-step statistical process that relies on both volumetric histogram processing and 2D segmentation. The completely unattended nature of such a procedure makes it possible to rapidly populate a large database of 3D point clouds describing healthy and unhealthy faces, enhancing the diagnosis of rare syndromes through statistical analyses.

Keywords: 3D histogram processing , 3D ultrasound , DICOM , Image segmentation

Digital documentation and high-quality 3D representation are always more requested in many disciplines and areas due to the large amount of technologies and data available for fast, detailed and quick documentation. This work aims to investigate the area of medium and small sized artefacts and presents a fast and low cost acquisition system that guarantees the creation of 3D models with an high level of detail, making the digitalization of cultural heritage a simply and fast procedure. The 3D models of the artefacts are created with the photogrammetric technique Structure From Motion that makes it possible to obtain, in addition to three-dimensional models, high-definition images for a deepened study and understanding of the artefacts. For the survey of small objects (only few centimetres) it is used a macro lens and the focus stacking, a photographic technique that consists in capturing a stack of images at different focus planes for each camera pose so that is possible to obtain a final image with a higher depth of field. The acquisition with focus stacking technique has been finally validated with an acquisition with laser triangulation scanner Minolta that demonstrates the validity compatible with the allowable error in relation to the expected precision.

Keywords: Cultural heritage , Focus stacking , Photogrammetry , Small artefacts , Structure from motion

The integration of underwater 3D data captured by acoustic and optical systems is a promising technique in various applications such as mapping or vehicle navigation. It allows for compensating the drawbacks of the low resolution of acoustic sensors and the limitations of optical sensors in bad visibility conditions. Aligning these data is a challenging problem, as it is hard to make a point-to-point correspondence. This paper presents a multi-sensor registration for the automatic integration of 3D data acquired from a stereovision system and a 3D acoustic camera in close-range acquisition. An appropriate rig has been used in the laboratory tests to determine the relative position between the two sensor frames. The experimental results show that our alignment approach, based on the acquisition of a rig in several poses, can be adopted to estimate the rigid transformation between the two heterogeneous sensors. A first estimation of the unknown geometric transformation is obtained by a registration of the two 3D point clouds, but it ends up to be strongly affected by noise and data dispersion. A robust and optimal estimation is obtained by a statistical processing of the transformations computed for each pose. The effectiveness of the method has been demonstrated in this first experimentation of the proposed 3D opto-acoustic camera.

Keywords: Optical and acoustic integration , Opto-acoustic vision , ROV navigation , Underwater 3D imaging

In this paper, a preliminary shadowgraph-based analysis of dust particles re-suspension due to loss of vacuum accident (LOVA) in ITER-like nuclear fusion reactors has been presented. Dust particles are produced through different mechanisms in nuclear fusion devices, one of the main issues is that dust particles are capable of being re-suspended in case of events such as LOVA. Shadowgraph is based on an expanded collimated beam of light emitted by a laser or a lamp that emits light transversely compared to the flow field direction. In the STARDUST facility, the dust moves in the flow, and it causes variations of refractive index that can be detected by using a CCD camera. The STARDUST fast camera setup allows to detect and to track dust particles moving in the vessel and then to obtain information about the velocity field of dust mobilized. In particular, the acquired images are processed such that per each frame the moving dust particles are detected by applying a background subtraction technique based on the mixture of Gaussian algorithm. The obtained foreground masks are eventually filtered with morphological operations. Finally, a multi-object tracking algorithm is used to track the detected particles along the experiment. For each particle, a Kalman filter-based tracker is applied; the particles dynamic is described by taking into account position, velocity, and acceleration as state variable. The results demonstrate that it is possible to obtain dust particles' velocity field during LOVA by automatically processing the data obtained with the shadowgraph approach.

The non-destructive evaluation of defects by automatic procedures is of great importance for structural components. Thanks to the developments of the non-contact ultrasonic techniques, the automation of the inspections is gaining a progressively important role. In this work, an automatic inspection technique for the evaluation of defects by the analysis of B-scan images obtained by a laser ultrasonic system is presented. The data are extracted directly from a B-scan map obtained for a panel with internal defects, and are used to build an image of the cross section of the panel. The proposed automatic procedure allows the definition of size, position and shape of defects in panels of known thickness.

Keywords: 2D defect shape definition , B-scan image analysis , Defect size definition , Laser UT system , NDE