Photometric Stereo Method Research Articles

Development of a design system for reverse engineering using the photometric stereo method

Development of a design system for reverse engineering using the photometric stereo method

Object surface recovery using a multi-light photometric stereo technique for non-Lambertian surfaces subject to shadows and specularities

This paper presents a new multi-light source photometric stereo system for reconstructing images of various characteristics of non-Lambertian rough surfaces with widely varying texture and specularity. Compared to the traditional three-light photometric stereo method, extra lights are employed using a hierarchical selection strategy to eliminate the effects of shadows and specularities, and to make the system more robust. We also show that six lights is the minimum needed in order to apply photometric stereo to the entire visible surface of any convex object. Experiments on synthetic and real scenes demonstrate that the proposed method can extract surface reflectance and orientation effectively, even in the presence of strong shadows and highlights. Hence, the method offers advantages in the recovery of dichromatic surfaces possessing rough texture or deeply relieved topographic features, with applications in reverse engineering and industrial surface inspection. Experimental results are presented in the paper.

Optical Measurement Of A 3D DisplacementField Using One Camera

The optical measurement of flat specimen deformation is a frequent task. In some applications it is necessary to measure both the in-plane and out-of-plane displacement or strain fields respectively. To achieve this goal a variety of methods have been developed in the past. This paper describes a novel technique based on the use of only one camera which employs a combination of two different optical methods to achieve 3D displacement measurements. The optical grid method for the measurement of the in-plane strain field and the proposed Coded Photometric Stereo method for the measurement of the out-of-plane displacement can be applied simultaneously. This approach reduces hardware part costs. The combined method is full field (all data acquired at once), non-contact and non-destructive.

On Estimation of the Luminance Distribution by the Diffuse Reflection Component of the Object Surface and Shape Measurement

Photometric stereo method can be used to find the orientation of the surface of an object to be measured. This method cannot avoid the influence of highlights in an image that are caused by the specular reflection components of mixed diffuse-specular objects. In a previous paper, we presented that estimate luminance distribution in highlight regions can eliminate influence of highlights in an image. However, the previous method is unstable when some highlight regions overlap. We have proposed a new method of measuring the shapes of mixed diffuse-specular objects. First, highlight is detected using the bi-directional reflectance distribution function. Next, highlight regions that overlap in a gray-level image are divided using the watershed algorithm. Then, the luminance distribution caused by the diffuse reflection components in the highlight regions is estimated based on the luminance distribution in a gray-level image without highlights. Shape measurement experiments using mixed diffuse-specular objects with different characteristic reflectances and shapes showed that our method can accurately measure the 3-D shape of mixed diffuse-specular objects.

Noise Reduction in Surface Reconstruction from a Given Gradient Field

We present a gradient space technique for noise reduction in surfaces reconstructed from a noisy gradient field. We first analyze the error sources in the recovered gradient field of a surface using a three-image photometric stereo method. Based on this analysis, we propose an additive noise model to describe the errors in the surface gradient estimates. We then use a vector space formulation and construct a multiscale orthonormal expansion for gradient fields. Using the sparse representation properties of this expansion, we develop techniques for reducing the gradient field noise by coefficient selection with thresholding. The simulation results indicate that the proposed technique provides significant improvement on the noise levels of both the estimated gradient fields and the reconstructed surfaces under heavy noise levels. Furthermore, the experiments using noisy photometric stereo image triplets of real range data suggest that the additive model remains viable after the nonlinear photometric stereo operation to provide accurate noise removal.

Measuring 3-D shapes of a human face using photometric stereo method with color light sources and slit patterns

We propose a new method for measuring 3-D facial shapes. This method uses two point light sources with red filters and a slit pattern projector with a blue filter. The time required for sampling data can be reduced by switching the color light rays at high speed, and hence a series of 3-D shapes of a human face at each instant can be measured. First, two facial images projected by both the red light rays and the blue light stripes are taken by the video camera, and these images are sepa- rated into two shading images and one slit image using the color infor- mation. Next, the normal vectors at points on the face are computed from the two shading images using the photometric stereo method. The 3-D coordinates of the points on the light stripes are measured from the slit image using the stereo vision algorithm. Finally, all the computed results are blended and the 3-D shape of the face is computed. Experimental results are shown to demonstrate the feasibility of the proposed method. © 2003 Society of Photo-Optical Instrumentation Engineers.

An improved method of photometric stereo using local shape from shading

This paper presents an improved photometric stereo (PS) method by integrating it with a local shape from shading (SFS) algorithm. PS produces the initial estimate of image for the global accuracy and also provides the recovery of albedo, SFS supplies the more detailed information within each homogeneous area. The quality of depth obtained by integrating PS and SFS is compared with the real depth using absolute dept error function, and the improvement ranging from 2.3 to 14% over PS is obtained.

Evaluating Fabric Wrinkle Degree with a Photometric Stereo Method

A method is proposed for fabric 3D surface reconstruction using a photometric stereo method. As an investigation of the efficacy of the method, the 3D surface of an AATCC standard wrinkle pattern is reconstructed and its wrinkle degree is measured using four index values that indicate the variation of the surface height value. The results show that there is a good linear correlation of the index value with the wrinkle degree of the pattern.

2322 ステレオ型イメージスキャナを用いた 3 次元形状復元

In this paper, we propose a shape reconstruction method combining a binocular stereo method and a photometric stereo method with three light sources. From a stereo images taken by a stereo image scanner, first, we estimate heights at edge points on the object by the binocular stereo method. Next, we obtain the albedo on the object surface near edge points by using estimated height. And we recover the object shape by the photometric stereo method with estimated albedos. In the experimental results, we show the effectiveness of the proposed method by recovering a cylindrical object shape from real stereo images.

Measuring and Modeling 3D Wrinkles in Fabrics

In this paper, we propose a new method for measuring fabric wrinkling based on integrating photometric stereo and image analysis techniques. In the traditional method. wrinkled fabrics are evaluated subjectively according to the grades of replica plates supplied by the American Association of Textile Chemists and Colorists (AATCC). But due to the disadvantages of subjective methods. which are tedious, time-consuming, subjective, and so on. image analysis techniques are used in this work as an objective method. However, serious problem occur when colored and patterned fabrics are tested with conventional images and processing techniques, although they work for tests of solid fabrics. In this paper, a photometric stereo method is used to extract three-dimensional surfaces of fabrics and successfully eliminate the influence of color and pattern. Image analysis techniques are furthered to extract wrinkling features of these reconstructed 3D images, and there is a good linear correlation between objective parameters and subjective AATCC grades.

Modeling the Luminance Distribution of 3-D Object Surface with Specular Reflection Component and Shape Measurement

A method is described for reducing the large error that often occurs in the measurement of three-dimensional (3D) objects using the photometric stereo method when there is a specular component in te reflection characteristics of the object’s surface. First, the change in the luminance distribution caused by the component is estimated based on the correlation between the modeled distribution and a gray-level image of the object, created by numerical experiment. Next, the shape measurement of the hemisphere object was carried out. Nurmerical experiment of three dimensional measurement of a hemispherical object showed that the mean error of normal vector can be reduced by as much as 90%. This method is applicable to machine vision, automatic manufacturing, and other industrial processes.

An Investigation of Fabric 3-D Surface Shape by Means of Photometric Stereo

A new method is proposed for fabric 3-D surface reconstruction by using a photometric-stereo technique. The idea of photometric stereo is to vary the direction of incident illumination between successive images while the viewing direction is kept constant. It is shown that this provides sufficient information to reconstruct a fabric-surface profile. In the proposed algorithm, the relative depth and its derivatives are obtained by using a finite-difference method. Synthetic images are first used to simulate the proposed photometric-stereo method, and then the actual fabric is used to test this method. The fabric-surface shape can be obtained from the 3-D contour data of the fabric. From the fabric-surface profile, we can extract a wrinkle-feature parameter. This is the base for analyzing and evaluating the degree of fabric wrinkle objectively.

Deformation detection on composite reflectance surfaces using two-image photometric stereo

We present a new two-image photometric stereo method that is capable of determining surface deformations when extended light sources are used and the imager and light sources are close to the surface. Surface reflectance is modeled using a linear combination of Lambertian, backscatter, and new extended-source Phong specular reflection component models. The ambiguity resulting from using only two light sources is resolved in the surface normal integration phase, in which the resulting surface is constrained to being a valid image-based triangular facet surface. The method has been tested on an application involving waste-storage drum inspection and yields good accuracy.

画像解析による非定常波高分布計測法の開発に関する研究

The ocean wave information is indispensable to grasp the ocean environment and also important to navigate ships economically and safety. Therefore, many research works about the way of measuring it have been done.Remote sensing technology using a satellite image is the useful way to measure it. This technology isn't suitable for getting the more detailed information of the limited part, though it is suitable for getting the average information of the large area.The projected light distribution method was kept proceeding to measure the detailed wave height distribution of the limited part. But, it's very difficult to exceed a laboratory scaleThis paper presents the way to measure the unsteady wave height distributions using the wave surface inclination distributions. The wave surface inclination distributions are analyzed images of the reflected light from the surface light source with the color distribution, by applying the Photometric stereo method. An unsteady wave height distribution is measured by this method and compared with the results measured by super-sonic wave height meters. The results show good agreement with them.

Topography and reflectance analysis of paper surfaces using a photometric stereo method

This paper describes a fast photometric stereo method, use- ful for the determination of surface-topography and reflectance. First we show how the irradiance in two images of a surface, illuminated from the left and right, respectively, can be used to determine the inclination of the surface elements. This requires a model for the light scattering proper- ties of the surface. The model used divides the scattered light into a bulk-scattered, Lambertian part and a surface-scattered, specular part. The specular part is eliminated with the help of crossed polarizers. The inclinations are integrated to give a surface height function with the help of a Wiener filter, which suppresses frequencies with an expected poor signal-to-noise ratio. Profiles measured with this method show good agreement with profiles obtained both with optical and mechanical scan- ning methods. The standard deviation between surface height measured with the different methods was about 1 mm. The ability of this method to study reflectance and topography at the same time makes it suitable for studies of the coupling between print result and surface topography. As an example, a gravure-printed paper surface with missing dots has been studied. It was found that missing dots occur at depressions, or pores, in the surface. © 2000 Society of Photo-Optical Instrumentation Engineers. (S0091-3286(00)00109-4)

3次元物体表面の反射光分布のモデル化と形状計測について

A method is described for reducing the large error that often occurs in the measurement of three-dimensional (3D) objects using the photometric stereo method when there is a specular component in the reflection characteristics of the object's surface.First, the change in the luminance distribution caused by the component is estimated based on the correlation between the modeled distribution and a gray-level image of the object, created by numerical experiment.Next, the shape measurement of the hemisphere object was carried out. Numerical experiment of a hemispherical object showed that the mean error of normal vector can be reduced by as much as 90%. Thismethod is applicable to machine vision, automatic manufacturing, and other industrial processes.

3D reconstruction of skin surface from photometric stereo images with specular reflection and interreflection

This paper describes a new method for reconstructing the structure of a skin surface replica from three shading images taken from three different lightings. Since the shading images include specular reflections and interreflection, the conventional photometric stereo method is not suitable for reconstructing its surface structure. In the proposed method, the evaluation function of the surface shape is defined, then the structure is reconstructed by optimizing the evaluation using simulated annealing. The experimental results from both synthetic images and real images demonstrate that the proposed method is more effective for shape reconstruction from shading images which include specular reflection and interreflection. © 1999 Scripta Technica, Electr Eng Jpn, 129(3): 51–58, 1999

A method for determining three‐dimensional surface orientation of objects with textures

AbstractTo reconstruct the shape of an object from its image, the conventional photometric stereo method requires a preprocessing (e.g., region segmentation based on gray levels) unless the reflectance of its surface is uniform and known; and the method is applicable only to a part of the surface illuminated by more than three light sources.This paper proposes a method based on the photometric stereo method, but with the newly introduced restriction conditions so that the method can be applied without preprocessing to an object with surface texture including regions illuminated by fewer than three light sources. The method has been applied successfully to two spherical objects: the mean square error in the reconstruction of an object where surface texture is 0.01723, and that without texture is 0.002063.

Gradient and curvature from the photometric-stereo method, including local confidence estimation

The photometric-stereo method is one technique for three-dimensional shape determination that has been implemented in a variety of experimental settings and that has produced consistently good results. The idea is to use intensity values recorded from multiple images obtained from the same viewpoint but under different conditions of illumination. The resulting radiometric constraint makes it possible to obtain local estimates of both surface orientation and surface curvature without requiring either global smoothness assumptions or prior image segmentation. Photometric stereo is moved one step closer to practical possibility by a description of an experimental setting in which surface gradient estimation is achieved on full-frame video data at near-video-frame rates (i.e., 15 Hz). The implementation uses commercially available hardware. Reflectance is modeled empirically with measurements obtained from a calibration sphere. Estimation of the gradient (p, q) requires only simple table lookup. Curvature estimation additionally uses the reflectance map R(p, q). The required lookup table and reflectance maps are derived during calibration. Because reflectance is modeled empirically, no prior physical model of the reflectance characteristics of the objects to be analyzed is assumed. At the same time, if a good physical model is available, it can be retrofitted to the method for implementation purposes. Photometric stereo is subject to error in the presence of cast shadows and interreflection. No purely local technique can succeed because these phenomena are inherently nonlocal. Nevertheless, it is demonstrated that one can exploit the redundancy in three-light-source photometric stereo to detect locally, in most cases, the presence of cast shadows and interreflection. Detection is facilitated by the explicit inclusion of a local confidence estimate in the lookup table used for gradient estimation.

Photometric stereo under a light source with arbitrary motion

A new photometric-stereo method for estimating the surface normal and the surface reflectance of objects without a priori knowledge of the light-source direction or the light-source intensity is proposed. First, I construct a p × f image data matrix I from p pixel image intensity data through f frames by moving a light source arbitrarily. Under the Lambertian assumption the image data matrix I can be written as the product of two matrices S and L, with S representing the surface normal and the surface reflectance and L representing the light-source direction and the light-source intensity. Using this formulation, I show that the image data matrix I is of rank 3. On the basis of this observation, I use a singular-value decomposition technique and useful constraints to factorize the image data matrix. This method can also be used to treat cast shadows and self-shadows without assumptions. The effectiveness of this method is demonstrated through performance analysis, laboratory experiment, and out-of-laboratory experiment.