Robert Sitnik

Digital fringe projection system for large-volume 360-deg shape measurement

We present a system for 3-D shape measurement in large volumes based on combined digital-fringe-Gray-code projection. With the help of a new calibration procedure, the system provides accurate results despite its crossed-axis configuration and unknown aberrations of the digital light projector and CCD camera. Also, the separate clouds of points captured from different directions are automatically merged into the main cloud. The system delivers results in the form of (x,y,z) coordinates of the object points with additional (R,G,B) color information about their texture. Applicability of the system is proven by presenting sample results of measurements performed on complex objects. The uncertainty of the system was estimated at 10 4 of the measurement volume.

On-line, collision-free positioning of a scanner during fully automated three-dimensional measurement of cultural heritage objects

Fully automated three-dimensional (3D) measurement (either with contact or non-contact measurement heads) cannot be easily completed without detailed a priori knowledge of the measured object. This requirement may be easily fulfilled for objects whose CAD models are available (for example, mechanical parts); however, for unique artifacts (such as sculptures), such models are usually unavailable. A description of the objects shape is required to create a list of positions at which the scanning head should be placed to completely measure the whole object (so-called next best views list). For unique objects, this analysis has to be performed on-line during measurements, using obtained data to update the list. For a fully automated measurement, the scanning system must be positioned mechanically, and the transition between subsequent head positions has to be collision free. This paper presents a novel system for 3D digitization of cultural heritage objects which allows one to perform completely automated shape measurements in a given volume with sampling resolutions up to 10000 points/ mm^2. Its main novelties are the next best view module for identifying the best position and orientation of the measurement head in subsequent directional scans (optimized for artifacts of size much bigger than the measurement volume) and flexible software modules for inverse kinematics and collision detection, which allow one to build a positioning system tailored to specific measurement objects. The system is used for digitization of a collection from the Wilanow Palace Museum, Warsaw.

New method of structure light measurement system calibration based on adaptive and effective evaluation of 3D-phase distribution

In the paper a new method for scaling of phase values into (x,y,z) co-ordinates supporting methods with absolute phase determination e.g. fringe projection / Gray code techniques is presented. It is based on calculation of characteristic polynomials describing relationships between phase values and real (x,y,z) co-ordinates in measurement volume. Coefficients of these polynomials are calculated on the base of phase distribution on known 2D-calibration model positioned manually in several unknown positions inside measurement volume. It introduces new way of calibration by two step process. First, it calculates exact positions of model on the base of its phase distributions and secondly it evaluates coefficient of polynomials. Applicability of the method described is proven by calibration of 3D shape measurement system based on unknown, commercially available projection and detection systems with unknown parameters of imaging optics and geometrical set-up. Exemplary measurement result of freeform object is presented.

Method of pectus excavatum measurement based on structured light technique

We present an automatic method for assessment of pectus excavatum severity based on an optical 3-D markerless shape measurement. A four-directional measurement system based on a structured light projection method is built to capture the shape of the body surface of the patients. The system setup is described and typical measurement parameters are given. The automated data analysis path is explained. Their main steps are: normalization of trunk model orientation, cutting the model into slices, analysis of each slice shape, selecting the proper slice for the assessment of pectus excavatum of the patient, and calculating its shape parameter. We develop a new shape parameter (I(3ds)) that shows high correlation with the computed tomography (CT) Haller index widely used for assessment of pectus excavatum. Clinical results and the evaluation of developed indexes are presented.

Automatic recognition of surface landmarks of anatomical structures of back and posture

Faulty postures, scoliosis and sagittal plane deformities should be detected as early as possible to apply preventive and treatment measures against major clinical consequences. To support documentation of the severity of deformity and diminish x-ray exposures, several solutions utilizing analysis of back surface topography data were introduced. A novel approach to automatic recognition and localization of anatomical landmarks of the human back is presented that may provide more repeatable results and speed up the whole procedure. The algorithm was designed as a two-step process involving a statistical model built upon expert knowledge and analysis of three-dimensional back surface shape data. Voronoi diagram is used to connect mean geometric relations, which provide a first approximation of the positions, with surface curvature distribution, which further guides the recognition process and gives final locations of landmarks. Positions obtained using the developed algorithms are validated with respect to accuracy of manual landmark indication by experts. Preliminary validation proved that the landmarks were localized correctly, with accuracy depending mostly on the characteristics of a given structure. It was concluded that recognition should mainly take into account the shape of the back surface, putting as little emphasis on the statistical approximation as possible.

3DMADMAC system: optical 3D shape acquisition and processing path for VR applications

This paper presents a full processing path from 3D-object scanning to its virtual representation with special focus on virtual museum application. Measurement technique is based on opto-numerical approach using structured light projection. The processing path consists of three main steps: filtering and preparing 3D-clouds of points, merging of directional measurements by local co-ordinate spaces unification and transforming from cloud of point representation to triangle mesh with photorealistic texture. The applicability of this path for objects with mixed diffusive-reflective surfaces and a very complex shape is shown on the example of results of a measurement session in the Kórnik Castle and our own 3D objects. The main goal of the session was 3D digitizing of a set cultural objects such as armours, bas-reliefs, sculptures and many more. The exemplary results of these full 3D models (including geometry and texture) scanning is presented.

Four-dimensional measurement by a single-frame structured light method

A four-dimensional [4D--three-dimensional (3D) shape varying in time] shape measurement system is described. A single 3D shape of an object is calculated from only one frame. The projected pattern is composed of sinusoidal intensity fringes and one color-encoded stripe, the analysis of which allows us to find the absolute coordinates of the measured object. During measurement, the position of the stripe changes due to the improvement of the quality of spatiotemporal unwrapping. The fringes deformed by the shape of the object are captured by a CCD camera and processed by an adaptive spatial carrier phase-shifting algorithm. The use of an algorithm based on fast Fourier transformation is proposed to approximate the local period of fringes. A new phase-unwrapping routine based on the spatiotemporal information is presented as well. All these features make the 3D shape measurement of an object in motion possible with the additional advantage of using a low-cost system. Experimental results of the developed method together with a preliminary assessment of measurement uncertainty are presented to show the validity of the method.

Archiving shape and appearance of cultural heritage objects using structured light projection and multispectral imaging

To create faithful reproduction of a cultural heritage object, it is crucial to gather information on intrinsic optical properties of the objects surface, as well as its geometry. An integrated device has been developed that performs a three-dimensional measurement using structured light projection, followed by multispectral imaging for precise color retrieval and directional illumination for estimating bidirectional reflectance distribution function (BRDF) parameters. The main advantage shown in this work is the use of only one detector during the whole acquisition process to assure ideal correspondence of multimodal surface data in the image space. A method is shown for performing the measurement using an integrated device. Methods of data organization and processing are described facilitating robust operation of the developed software. A prototype setup for the integrated system is presented together with measurement parameters and sample measurement.

Integrated three-dimensional shape and reflection properties measurement system

Creating accurate three-dimensional (3D) digitalized models of cultural heritage objects requires that information about surface geometry be integrated with measurements of other material properties like color and reflectance. Up until now, these measurements have been performed in laboratories using manually integrated (subjective) data analyses. We describe an out-of-laboratory bidirectional reflectance distribution function (BRDF) and 3D shape measurement system that implements shape and BRDF measurement in a single setup with BRDF uncertainty evaluation. The setup aligns spatial data with the angular reflectance distribution, yielding a better estimation of the surfaces reflective properties by integrating these two modality measurements into one setup using a single detector. This approach provides a better picture of an objects intrinsic material features, which in turn produces a higher-quality digitalized model reconstruction. Furthermore, this system simplifies the data processing by combining structured light projection and photometric stereo. The results of our method of data analysis describe the diffusive and specular attributes corresponding to every measured geometric point and can be used to render intricate 3D models in an arbitrarily illuminated scene.

New concept of fast hybrid contact and no-contact measurement for automotive industry

In this paper a new concept of a hybrid Opto-Mechanical Measurement Machine (OMMM) is presented. This concept combines advantages of the high accuracy contact CMM and the high measurement speed non-contact optical techniques. OMMM is developed for measurement of large-size elements, especially for the automotive industry. It introduces high-speed and high-accuracy measurement procedure which is performed in four steps: 1) optical structured-light measurement of the entire surface, 2) analysis of model parts, determination which area should be re-measured with higher accuracy, 3) measurement with a CMM, 4) final metrological data analysis. Data gathered by such a system are compared with the CAD model of the object, thus allowing automatic error control of every manufactured product.