Paul Besl

Intel · Software and Solutions Group (SSG)

This chapter examines the hStreams library which supports programming of a heterogeneous system by abstracting programming to be akin to feeding a system with streams of actions (computations, data transfers, and synchronizations). Use of the library is illustrated with some problems from the field of linear algebra, endeavoring to show the speed a...

A heterogeneous system offers differing compute capabilities which need to be given work to do, fed data, and synchronized. The hStreams library offers an abstraction that controls these capabilities of a system with "streams of actions" made up of computations, data transfers and synchronizations. This simple FIFO stream abstraction is very powerf...

http://deepblue.lib.umich.edu/bitstream/2027.42/3570/5/anf5941.0001.001.pdf http://deepblue.lib.umich.edu/bitstream/2027.42/3570/4/anf5941.0001.001.txt

http://deepblue.lib.umich.edu/bitstream/2027.42/3567/5/anf5939.0001.001.pdf http://deepblue.lib.umich.edu/bitstream/2027.42/3567/4/anf5939.0001.001.txt

The perception of surfaces plays a key role in three-dimensional (3-D) object recognition. 3-D objects in 2-D images are recognized by the matching of perceived surface descriptions with stored object models, and the first step is image segmentation. An approach for segmenting images into symbolic surface descriptions is discussed. The approach may...

The sections in this article are

Some systems provide 2D environments where users can (1) import imagery from 2D cameras, (2) create new geometry and imagery, (3) combine the imported with the new, and (4) create high-quality production output. In contrast, most 3D modeling systems provide either non-uniform [knot sequence] rational b-splines (NURBS) or polygonal modeling with no...

Current methods for extracting surface-models from “clouds of points” use standard CAGD techniques and guarantee neither fairness nor closeness of fit — important requirements for applications like Aesthetic Surface Design (Styling) and Reverse Enginnering. This paper elaborates on a new approach, combining CAGD with Image Processing, to derive sur...

Existing techniques for approximating discrete data with smooth surfaces arc based on either curve fitting or edge detection. Furthermore, they often rely on the user’s expertise for determining tht, number of polynomial segments included in each surface. This paper describes :i technique for approximating 3-D points with a functional surfiace z =...

Triangles have often been overlooked as a serious generalpurpose geometric representation for many applications owing to storage, accuracy, and combinatoric concerns. This paper advocates triangles as a very workable common representation for almost all types of 3D surface data and for almost all applications, including 3D computer vision, by exami...

This paper describes a general purpose, representation independent method for the accurate and computationally efficient registration of 3-D shapes including free-form curves and surfaces. The method handles the full six-degrees of freedom and is based on the iterative closest point (ICP) algorithm, which requires only a procedure to find the close...

The authors describe a general-purpose, representation-independent method for the accurate and computationally efficient registration of 3-D shapes including free-form curves and surfaces. The method handles the full six degrees of freedom and is based on the iterative closest point (ICP) algorithm, which requires only a procedure to find the close...

The authors deal with the principal patch representation of visible surfaces in a scene, an algorithm for creating a stable surface model in this representation, and an algorithm for extending this model into subsequent views. The representation is based on the fact that the lines of minimum and maximum curvature on a generic smooth surface are ort...

A wide variety of sensing techniques allow the direct measurement of the three-dimensional (3-D) coordinates of closely spaced points in a scene. A optical profilometer or a single light-stripe range sensor measures a one-dimensional depth profile z i = f(x i ) of a surface along a line specified by a y value. A single-view range imaging sensor [Be...

It is a common practice in computer vision and image processing to convolve rectangular constant coefficient windows with digital images to perform local smoothing and derivative estimation for edge detection and other purposes. If all data points in each image window belong to the same statistical population, this practice is reasonable and fast....

Active, optical range imaging systems collect three-dimensional coordinate data from object surfaces. These systems can be useful in a wide variety of automation applications, including shape acquisition, bin picking, assembly, inspection, gauging, robot navigation, medical diagnosis, cartography, and military tasks. The range-imaging sensors in su...

Geometric matching algorithms and geometric representations are examined for point sets, curves, surfaces, volumes, and their space-time trajectories. The author focuses mainly on general-purpose techniques. Geometric matching benefits greatly from taking advantage of symmetries and special features, e.g. dihedral edges and circles or any three rel...

Active, optical range imaging sensors collect three-dimensional coordinate data from object surfaces and can be useful in a wide variety of automation appli- cations, including shape acquisition, bin picking, assem- bly, inspection, gaging, robot navigation, medical diagno- sis, and cartography. They are unique imaging devices in that the image dat...

The solution of the segmentation problem requires a mechanism for partitioning the image array into low-level entities based on a model of the underlying image structure. A piecewise-smooth surface model for image data that possesses surface coherence properties is used to develop an algorithm that simultaneously segments a large class of images in...

The task of automating the visual inspection of pin-in-hole solder joints is addressed. Two approaches are explored: statistical pattern recognition and expert systems. An objective dimensionality-reduction method is used to enhance the performance of traditional statistical pattern recognition approaches by decorrelating feature data, generating f...

Not Available

Computing surface characteristics for pixel grouping purposes is the first step toward the final goals of object recognition and image understanding. The range image object recognition problem is being addressed with the aim of developing a general purpose approach that handles arbitrary surface shapes and arbitrary viewing directions. To handle th...

An analytical framework for object recognition and image segmentation was formulated in Chapter 2. It was proposed that objects can be recognized in range images in a general manner through surface characterization, surface segmentation, and surface matching. Although surface matching is inherently model-dependent, it was proposed that surface char...

Both the ability to organize signal samples into symbolic primitives and the ability to recognize groups of symbolic primitives are necessary for machine perception. Just as a person must identify phonemes to form words to understand a sentence, a computer vision system must identify 3-D surfaces belonging to 3-D objects to understand a 3-D scene....

The iterative region growing algorithm isolates surface regions sequentially based on the original HK-sign map regions. As discussed in Chapter 4, three different region descriptions are recorded for each region so that the effects of the sequential processing might be accounted for in the final segmentation. The static first-within-tolerance regio...

Edges play a dominant role in computer vision research, but have received little attention so far. Although independent edge detection and edge linking algorithms are not of direct concern to the digital surface segmentation algorithm, the topic of parametric edge description is of significant interest. Since the approximate shape of smooth graph s...

This chapter critiques and displays the results of the digital surface segmentation algorithm applied to a set of twenty-two test images containing real and noisy synthetic range images and real intensity images. These images were selected from a group of images used as a test image database for the experiments. All input images are 128×128 pixels...

The perception of surface geometry plays a key role in 3-D object recognition. It is proposed that general-purpose recognition of 3-D objects from 2-D images can be performed without complete reliance on special purpose features by matching perceived surface descriptions with stored models of objects. In range imagery, the perceived surface descrip...

http://deepblue.lib.umich.edu/bitstream/2027.42/3565/5/anf5969.0001.001.pdf http://deepblue.lib.umich.edu/bitstream/2027.42/3565/4/anf5969.0001.001.txt

In recent years there has been a tremendous increase in computer vision research using range images (or depth maps) as sensor input data. The most attractive feature of range images is the explicitness of the surface information. Many industrial and navigational robotic tasks will be more easily accomplished if such explicit depth information can b...

Perception of surfaces plays a fundamental role in three-dimensional object recognition and image underst and ing. A range image explicitly represents the surfaces of objects in a given field of view as an array of depth values. Previous research in range image underst and ing has limited itself to extensions of edge-based intensity image analysis...

In recent years there has been a tremendous increase in computer vision research using range images (or depth maps) as sensor input data. The most attractive feature of range images is the explicitness of the surface information. Many industrial and navigational robotic tasks will be more easily accomplished if such explicit depth information can b...

An approach is described for the automatic inspection of solder joints on printed circuit boards. Common defects are identified in solder joints and a joint is classified as being good or belonging to one of the defective classes. The motivation for this classification is not just the detection of defective joints, but the desire to automatically t...

This paper describes an approach for automatic inspection of solder joints on printed circuit boards using gray-scale images. Common defects in solder joints are recognized using features computed from segmented solder joint subimages. Unacceptable joints are assigned to one of several defective classes. Defect classification, rather than just dete...

A general purpose computer vision system must be capable of recognizing three-dimensional (3-D) objects. This paper proposes a precise definition of the 3-D object recognition problem, discusses basic concepts associated with this problem, and reviews the relevant literature. Because range images (or depth maps) are often used as sensor input inste...

http://deepblue.lib.umich.edu/bitstream/2027.42/3569/5/anf5944.0001.001.pdf http://deepblue.lib.umich.edu/bitstream/2027.42/3569/4/anf5944.0001.001.txt

A direct construction approximation algorithm is proposed for fitting B-spline curves and surfaces to polylines and quadmeshes, respectively. This iterative refinement algorithm has the following properties for N data points: it (a) uses O(N) memory, (b) can execute in roughly O(NlogN) time, (c) computes non-uniformly spaced knot and control point...

Perception of surfaces plays a fundamental role in three-dimensional object recognition and image understanding. A range image explicitly represents the surfaces of objects in a given field of view as an array of depth values. Previous research in range image understanding has limited itself to extensions of edge-based intensity image analysis or t...