Principal components analysis and neural network implementation of photometric stereo

Iwahori, Yuji; Woodham, Robert J.; Bagheri, A.

doi:10.1109/pbmcv.1995.514676

Cited by 23 publications

(4 citation statements)

References 8 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Here, an explanation is added about the NN architecture (Iwahori, Woodham, & Bagheri, 1995). NN seeks a representative vector of each cluster through a Gaussian kernel in a multidimensional space based on a learning sample.…”

Section: Shape Recovery Through Optimization and Nn Learningmentioning

confidence: 99%

“…However, there may be a problem such that biased learning is done based on the selected data. Paper (Iwahori et al, 1995) corresponds to this problem by using Orthogonal Least Squares (OLS) method which applies the regression model. Neurons are added with one by one to the NN until the sum of square errors reaches less than the target error or number of neurons becomes over the limitation, that is, number of learning epochs.…”

Section: Shape Recovery Through Optimization and Nn Learningmentioning

confidence: 99%

See 1 more Smart Citation

Recovering Polyp Shape from an Endoscope Image Using Two Light Sources

Iwahori

Kijsirikul

Usami

et al. 2017

International Journal of Software Innovation

View full text Add to dashboard Cite

This paper proposes a new approach to recover the polyp shape from an endoscope image using a photometric constraint equation considering two light sources. The procedures are as follows. First, obtain the initial depth distributions by optimizing photometric equation obtained from two light sources. Next, obtain the surface normal vector from depth using numerical difference at each point. Then the mapping between the obtained normal vector and true normal vector is learned using Radial Basis Function Neural Network for a Lambertian sphere, and learning is generalized to another actual polyp image. Finally, optimize the depth using the obtained surface normal to recover the final 3D shape. The validity is confirmed of this method in comparison with the previous methods via computer simulation and experiments using actual endoscope images.

show abstract

Section: Shape Recovery Through Optimization and Nn Learningmentioning

confidence: 99%

Section: Shape Recovery Through Optimization and Nn Learningmentioning

confidence: 99%

Recovering Polyp Shape from an Endoscope Image Using Two Light Sources

Iwahori

Kijsirikul

Usami

et al. 2017

International Journal of Software Innovation

View full text Add to dashboard Cite

show abstract

“…Chen et al [2] recover the albedo values for color images using photometric stereo. In [3][4][5], the authors use a calibrating object of known shape and constant albedo to establish a nonlinear mapping between the image irradiance and shape of the object in the form of a lookup table. For photometric stereo, a neural network-based approach is presented in [6] for a rotationally symmetric object with nonuniform reflectance.…”

Section: Introductionmentioning

confidence: 99%

A Practical Approach for Simultaneous Estimation of Light Source Position, Scene Structure, and Blind Restoration Using Photometric Observations

Sharma

Joshi

2008

EURASIP J. Adv. Signal Process.

View full text Add to dashboard Cite

Given blurred observations of a stationary scene captured using a static camera but with different and unknown light source positions, we estimate the light source positions and scene structure (surface gradients) and perform blind image restoration. The images are restored using the estimated light source positions, surface gradients, and albedo. The surface of the object is assumed to be Lambertian. We first propose a simple approach to obtain a rough estimate of the light source position from a single image using the shading information which does not use any calibration or initialization. We model the prior information for the scene structure as a separate Markov random field (MRF) with discontinuity preservation, and the blur function is modeled as Gaussian. A proper regularization approach is then used to estimate the light source position, scene structure, and blur parameter. The optimization is carried out using the graph cuts approach. The advantage of the proposed approach is that its time complexity is much less as compared to other approaches that use global optimization techniques such as simulated annealing. Reducing the time complexity is crucial in many of the practical vision problems. Results of experimentation on both synthetic and real images are presented.

show abstract

“…It has been shown that the extraction of surface normal orientation and local curvature for such objects is possible through the use of multiple images of the same scene taken from the same viewpoint but under different known, precisely-calibrated illumination conditions [4,5,6,11,12].…”

Section: Introductionmentioning

confidence: 99%

Gaussian curvature from photometric scatter plots

Angelopoulou

Proceedings Workshop on Photometric Modeling for Computer Vision and Graphics (Cat. No.PR00271)

View full text Add to dashboard Cite

Local surface curvature is an important shape descriptor, especially for smooth featureless objects. For this family of objects, if their surface is matte, there is a one-to-one mapping between their surface normal map and the photometric data collected from a scene under three different illumination conditions. This mapping allows for the extraction of the sign and the magnitude of Gaussian curvature (to within a constant multiple) directly from intensity values. Because all the computations are performed in photometric space, the normal map is never recovered. This implies that the precise location of the light sources is not needed for any of the computations. Experiments show that a simple setup with minimal illumination planning and calibration is sufficient for the extraction of Gaussian curvature for smooth diffuse surfaces.

show abstract

Principal components analysis and neural network implementation of photometric stereo

Cited by 23 publications

References 8 publications

Recovering Polyp Shape from an Endoscope Image Using Two Light Sources

Recovering Polyp Shape from an Endoscope Image Using Two Light Sources

A Practical Approach for Simultaneous Estimation of Light Source Position, Scene Structure, and Blind Restoration Using Photometric Observations

Gaussian curvature from photometric scatter plots

Contact Info

Product

Resources

About