S. Quinlan. Efficient distance computation between non-convex objects. In Proceedings of IEEE International Conference on Robotics and Automation, volume 4, pages 3324--3329, 1994.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....for performing proximity queries between general polyhedra. Specifically, sphere trees, cone trees, axis aligned box trees, oriented box trees, k d trees and octrees, trees based on S bounds, and k dops have been used for fast intersection queries for general polyhedra, as well as polygon soups [13, 20, 10, 21, 1, 14]. For most scenarios, these hierarchies excel at intersection detection but do not do so well when it comes to distance computation. 2.3 Multiresolution Techniques Multiresolution modeling techniques, such as model simplification, have been proposed to extract the shape of the underlying ....

S. Quinlan. Efficient distance computation between non-convex objects. In Proceedings of International Conference on Robotics and Automation, pages 3324--3329, 1994.

.... The most popular ones are hierarchical decomposition algorithms, which precompute a multi level bounding approximation of every object in an environment, using primitive volumes such as spheres, axis aligned bounding boxes, or oriented bounding boxes [CLMP95, GLM96, Hub96, KHM 98, KPLM98, Qui94] Experiments reported in [SA01] indicate that the PQP package [GLM96] tests two objects, described by 500,000 triangles each, in times ranging between 0.0001 and 0.0025 seconds (on an Intel Pentium III 1GHz processor) depending on the actual distance between the two objects. 6 The use of ....

S. Quinlan. Efficient distance computation between non-convex objects. In Proc. IEEE Int. Conf. on Robotics and Automation, pages 3324--3329, 1994.

....collision detection algorithms have required a large amount of geometrical intersection tests. To improve the efficiency of such algorithms, hierarchical representations of entities were developed to localise the areas where the actual collision occurred. These include sphere trees [14] 21][22], OBB trees (Oriented bounding boxes) 11] and hierarchies of k DOPs (Discrete Orientation Polytopes) 17] While speed and efficiency has been the main focus of such research, the issue of a constant frame rate is also paramount. This problem has been addressed in part by exploiting coherence ....

Quinlan, S. Efficient Distance Computation between Non-Convex Object. Proceedings International Conference on Robotics and Automation. 3324-3329. 1994.

....a complex combination of results that extend from considering mechanical parameters to solve great systems of differential equations, to detect collisions, to model reactions to the collisions, among others. These processes generate problems when the simulations must be executed in real time [3, 5, 10, 11, 12, 17]. On the other hand the design of polyhedral objects of free 3D form is in general an arduous task. The constructive geometric solid techniques are very useful in the context of industrial design, but they seem to be less flexible for applications with animation of characters or simulation of ....

S. Quinlan. "Efficient distance computation between non convex objects". In Proceedings of the IEEE International Conference On Robotics and Automation, 1994, 3324--3329.

....data structure is simple to produce automatically, and lends itself to efficient and elegant recursive algorithms. The disadvantage of this approach is that each level of the hierarchy does not fit the underlying object very tightly. Sphere Trees [Hubbard 1995a, 1996] Palmer and Grimsdale 1995][Quinlan 1994]. The main advantages of using spheres are that they are rotationally invariant, making them very fast to update, and it is very simple to test for distances between them, and test for overlaps. The disadvantage is that spheres do not approximate certain types of objects very efficiently. Hubbard ....

Quinlan, S. (1994) Efficient Distance Comp utation between Non-Convex Object. Proceedings International Conference on Robotics and Automation. 3324-3329.

....Set p to be the midpoint of line segment xx . 7. if CLEARANCE(p) returns TRUE then 8. Insert p into G as a new milestone. To perform the bridge test, RBB uses only one geometric primitive, CLEARANCE, which can be implemented very efficiently using a collision detection algorithm (see, e.g. Qui94, GLM96] The bridge test is purely local and does not require processing the global geometry of C. Choosing the probability density . The density determines how frequently a bridge of a particular length and orientation is chosen. We prefer short bridges over longer ones in order to increase ....

S. Quinlan. Efficient distance computation between non-convex objects. In Proc. IEEE Int. Conf. on Robotics and Automation, pages 3324--3329, 1994.

....NO PATH answer. The cost of computing an explicit representation of F in a high dimensional space is prohibitive. Instead, a PRM uses the implicit representation of F provided by ds, which only computes distances in 5 2 or 5 a and admits several reasonably efficient implementa tions (e.g. [7, 8, 12, 18, 19, 20, 22, 25]) In this paper we often use simple illustrative examples of 2 D free spaces, which could easily be handled by other planning techniques. Keep in mind that in practical problems, it is often not realistic to explicitly represent F. 3 Other Sampling Strategies The planner in [5] uses a partially ....

....obstacles. To be more specific, let the robot be made of p rigid bodies L, Lp. Let ti be the shortest translation that Li must undergo be fore it touches an obstacle. The distance dist(q) is min= ti . As we mentioned in Section 2, a num ber of techniques have been proposed to implement dist [7, 8, 12, 18, 19, 20, 22, 25]. To be usable by new roadmap, the function dist(q) should be extended to compute the (negated) penetra tion distance of the robot into the obstacles at configurations where q F. Then, q G F if and only if clst(q) 5. Consistently with the above definition of clst(q) when q G F, we can define ....

S. Quinlan. Efficient Distance Computation Between Non-Convex Objects. Proc. IEEE fnt. Conf. on Rob. and Aut., San Diego, CA, pp. 3324-3330, 1994.

....collaboration over the Internet in latency dependent or latency moderated modes. Many surgical and non surgical virtual tools have been created and their interactions with tissue have been implemented. Collision detection is provided through an enhanced[19] bounding sphere algorithm[20]. In addition, extra features such as voice input output, real time texture mapped video input, stereo and head mounted display support, and replicated display facilities are implemented. 2.1 Architecture An overview of the architecture is provided in Figure 1. It consists of a main program ....

....Later, static partition methods, Axis Aligned Bounding Boxes[24] and Oriented Bounding Boxes[25] were implemented. While these methods worked well in many cases, a more general scheme that better supported deformable objects was sought. For this reason, we moved to a Bounding Sphere algorithm[20], with enhancements for deformable objects[19] The generality of this method, and its fast update capability, provided a reasonable tradeoff for many cases. As with any hierarchical method, ultimately the detection method decomposes to testing collisions between primitive elements, such as ....

Quinlan, S, "Efficient Distance Computation Between Nonconvex Objects", Proc. IEEE Int Conf on Robotics and Automation, pp. 3324-3329, 1994.

....haptic rates ( 1000Hz) The system uses constraints to limit processing requirements to provide for increased scalability while maintaining adequate update rates. Collision Detection Real time, haptic rate collision detection[6] was performed by a modified hierarchical Bounding Spheres algorithm[7]. Modifications to this algorithm were made to support rapid update of the bounding sphere tree for deformable objects[8] as well as a number of other enhancements to bound the regions of detection and update to increase performance to haptic rates. Instruments Interactions Several different ....

Quinlan, S, "Efficient Distance Computation Between Nonconvex Objects", Proc. IEEE Int Conf on Robotics and Automation, pp. 3324-3329, 1994.

....separation distances have been extensively researched. Many are restricted to convex objects [Cameron 97, Ehmann01, Gilbert88, Lin91, Mirtich98] or are based on hierarchical bounding volume or spatial data structures that require considerable precomputation and are best suited for rigid geometry [Hubbard93, Quinlan94, Gottschalk96, Johnson98, Klosowski98]. Some algorithms handle dynamically deforming geometry by assuming that motion is expressed as a closed form function of time [Snyder93] or by using very specialized algorithms [Baraft92] In our approach, we emphasize the handling of non convex, dynamically deformable objects with no ....

....the complexity of the algorithm. We use coarse fixed height bounding volume hierarchies to achieve this balance between speed and complexity, and between CPU and graphics usage. There are many general and efficient algorithms available for localizing geometry based on bounding volume hierarchies [Gottschalk96, Hubbard93, Johnson98, Quinlan94]. However, for exact collision detection these algorithms typically perform well only on static geometry where the hierarchy can be precomputed. In order to handle dynamic deformable geometry with no precomputation, we use coarse levels for efficient trivial rejection and obtain reasonable ....

S. Quinlan, Efficient Distance Computation between Non-Convex Objects. International Conf. on Robotics and Automation, 3324-3329, 1994

....much more than 90 ) Several approaches are possible to reduce the overall cost of collision checking: # Design faster collision checking algorithms. However, efficient techniques already exist, e.g. some checkers pre compute a hierarchy of bounding volumes for every object in an environment [8, 9, 10]. For each collision query, they then use the hierarchies to quickly rule out large portions of objects that cannot collide. The scale up well to environments where object surfaces are described by several 100,000 triangles [6] # Design smarter sampling strategies. For example, the strategy in ....

S. Quinlan. Efficient distance computation between non-convex objects. In Proc. IEEE Int. Conf. Rob. & Autom., pages 3324--3329, 1994.

....of nodes with no significant performance penalty. 3. Collision Detection Research on collision detection between rigid objects has a long history in robotics, graphics, and solid modeling. Two main families of methods have been proposed: feature based (e.g. 1, 17, 18] and hierarchical (e.g. [6, 9, 14, 20, 23]) A feature based method exploits temporal and spatial coherence in the geometric model to maintain the pair of closest features. A hierarchical method pre computes a hierarchy of bounding volumes for every object. During a collision test, the hierarchies are used to quickly discard large subsets ....

....query would be too time consuming. Below, we propose an algorithm that does not modify the topology of the hierarchy representing an object, but only updates the size and location of the primitive volumes labeling the nodes of this hierarchy. Our algorithm derives from the one proposed by Quinlan [23] for rigid objects. Fewer works exist for deformable objects (e.g. 2, 26, 11, 24, 27] The algorithm in [26] is the closest to ours. It also builds a tree of fixed topological structure. But this tree is not balanced (which may seriously increase the cost of collision queries) and its nodes are ....

[Article contains additional citation context not shown here]

S. Quinlan. Efficient distance computation between nonconvex objects. In Proceedings of the IEEE International Conference On Robotics and Automation, pages 3324--3329, 1994.

....partition the model into a hierarchical spatial bounding volume. Nodes of the hierarchy may be pruned by obtaining a lower bound on the distance to the contained geometry and comparing that distance to an upper bound on the minimum distance obtained through a depth first descent in the hierarchy [12,8] or a sample point on the surface [11] The primary research thrust in this area has focused on methods for quickly determining the lower bound on the distance at a node. A different approach has been found in work on sculptured surfaces, such as B splines [13,10] Often a local solution to the ....

#Quinlan, Sean. Efficient Distance Computation between Non-Convex Objects, IEEE Int. Conference on Robotics and Automation, pp. 3324-3329, 1994.

....choice of a BV type. In particular, there is a trade off between the speed of BV operations and the tightness of fit which affects the number of operations performed. Most of the earlier work in this area was based on using simple BV types like spheres and AABBs [CLMP95, Hub93, BKSS90, HKM95, Qui94] In the last few years, many researchers have investigated the use of tight fitting BVs for proximity queries. These include using of OBBs [GLM96, BCG 96] spherical shells [KPLM98] and kDOPs [HKM96, KHM 98] for collision detection. For many close proximity configurations, these tight ....

....and kDOPs [HKM96, KHM 98] for collision detection. For many close proximity configurations, these tight fitting BVs result in improved performance as compared to BVHs composed of spheres or AABBs. One of the commonly used algorithm for distance computation algorithm uses spheres as a BV type [Qui94] Johnson and Cohen [JC98] have used OBBs for distance computation as well. 1.1 Main Contributions In this paper, we present a new distance computation algorithm that uses hierarchies of Rectangle Swept Spheres (RSS) This BV corresponds to the set of points obtained by sweeping the center of a ....

[Article contains additional citation context not shown here]

S. Quinlan. Efficient distance computation between non-convex objects. In Proceedings of International Conference on Robotics and Automation, pages 3324--3329, 1994.

....at most a small fraction of the underlying primitives. In our approach a hierarchical bounding representation for the object is constructed to take advantage of the spatial coherence inherent in most objects. The bounding representation, based on spheres, is similar to that proposed by Quinlan[20]. 6.3 Low Level Control A virtual proxy reduces the task of the low level servo controller to minimization of the error between the configuration of the proxy and position of the haptic device. In our current implementation a simple operational space proportional derivative (PD) controller ....

Quinlan, S., "Efficient Distance Computation between NonConvex Objects," Int. Conf. of Rob. and Automation , (April 1994).

....its corresponding vertex in the other configuration. Although there are many other types of metrics, such as Riemannian metrics [17] Hausdorff distance [9] rotation distance [20] growth distance [15] or the minimum distance between any point on the robot in the two configurations (e.g. 8] [16]) they were not selected because they were either too expensive for use in prms or were only applicable to convex objects. III. Local Planners Local planners are used in prms to make connections between nodes when building the roadmap. They must be fast (so many connections can be attempted) ....

S. Quinlan. Efficient distance computation between non-convex objects. In Proc. IEEE Int. Conf. Robot. Autom. (ICRA), pages 3324--3330, 1994.

....the primitives (and consequently the bounding volumes) are rotated and translated in the scene. Unfortunately, as recognized in [40] these objectives are usually in conflict, so a balance among them must be reached. Commonly used object partitioning representations use hierarchies of spheres [26,56,69,65,43] or spherical shells [8,50] for bounding at different resolution levels. Inner and outer bounds are often used. Volume bounding strategies can be used in conjunction with boundary representations. Hierarchies of volumes enclosing boundary features permit focussing on those susceptible 18 of ....

Quinlan, S. Efficient distance computation between non-convex objects. In Proc. IEEE Int. Conf. on Robotics and Automation (San Diego (CA), 1994), vol. 4, pp. 3324--3329.

....partitioning or frame to frame coherence, there is relatively little available on hybrid approaches combining two or more such techniques. Bounding Volume Hierarchies: A number of algorithms based on hierarchical representations have been proposed. The set of bounding volumes include spheres [17, 29], axis aligned bounding boxes [5, 16] oriented bounding boxes [15, 4] approximation hierarchies based on S bounds [7] spherical shells [20] and k dop s [19] In the close proximity scenarios, hierarchies of oriented bounding boxes (OBBTrees) appear superior to many other bounding volumes. ....

....also has a collision detection library probably using BSPTrees [31, 33] Ruspini et al. 30] have presented a haptic interface library HL that uses a multi level control system to effectively simulate contacts with virtual environments. It uses a bounding volume hierarchy based on sphere trees [29]. Nahvi et al. 25] have designed a haptic display system for manipulating virtual mechanisms derived from a mechanical CAD design. It uses the Sarcos Dexterous Arm Master and Utah s Alpha 1 CAD system, with algorithmic support from a tracing algorithm and a minimum distance framework developed ....

S. Quinlan. Efficient distance computation between nonconvex objects. In Proceedings of International Conference on Robotics and Automation, pages 3324--3329, 1994.

....Such a data structure is simple to produce automatically, and lends itself to efficient and elegant recursive algorithms. The disadvantage of this approach is that each level of the hierarchy does not fit the underlying object very tightly. Sphere Trees [Hubbard 1996] Palmer and Grimsdale 1995][Quinlan 1994]. The main advantages of using spheres are that they are rotationally invariant, making them very fast to update, and it is very simple to test for distances between them, and test for overlaps. The disadvantage is that spheres do not approximate certain types of objects very efficiently. Hubbard ....

Quinlan, S. (1994) Efficient Distance Computation between Non-Convex Object. Proceedings International Conference on Robotics and Automation. 3324-3329.

....in [8] runs in roughly constant time even when the closest features change. However, they are not applicable in the case of a surgery simulator, since organs are generally non convex, and deform over time. Among the collision detection methods that are applicable to more general polygonal models [10, 2, 12, 4, 11, 13, 5, 7], almost all of the optimizations rely on a pre computed hierarchy of bounding volumes. The solutions range from axis aligned box trees, sphere trees [12, 11, 7] or BSP trees, to more specific data structures [2] All these techniques, which perform very efficient rejection tests, may ....

....and deform over time. Among the collision detection methods that are applicable to more general polygonal models [10, 2, 12, 4, 11, 13, 5, 7] almost all of the optimizations rely on a pre computed hierarchy of bounding volumes. The solutions range from axis aligned box trees, sphere trees [12, 11, 7], or BSP trees, to more specific data structures [2] All these techniques, which perform very efficient rejection tests, may considerably slow down when objects are very close, i.e. when the bounding volumes have multiple intersections. Among the recent approaches for finding bounding volumes that ....

S. Quinlan. Efficient distance computation between nonconvex objects. In International Conference of Robotics and Automation, pages 3324--3329, 1994.

....collision detection algorithms have required a large amount of geometrical intersection tests. To improve the efficiency of such algorithms, hierarchical representations of entities were developed to localise the areas where the actual collision occurred. These include sphere trees [14] 20][21], OBB trees (Oriented bounding boxes) 11] and hierarchies of k DOPs (Discrete Orientation Polytopes) 17] While speed and efficiency has been the main focus of such research, the issue of a constant frame rate is also paramount. This problem has been addressed in part by exploiting coherence ....

Quinlan, S. (1994) Efficient Distance Computation between NonConvex Object. Proceedings International Conference on Robotics and Automation. 3324-3329.

....collisions, intersections, and minimum separation distances have been extensively researched. Many are restricted to convex objects [2,4,6,16] or are based on hierarchical bounding volume or spatial data structures that require considerable precomputation and are best suited for rigid geometry [8,12,14,19]. Some algorithms handle dynamically deforming geometry by either having prior knowledge of motion trajectories [22] or by using very specialized algorithms [1] In our approach, we emphasize the handling of non convex, dynamically deformable objects with no precomputation or knowledge of object ....

....field. The lengths represent the distance to the boundary (right) The top level bounding boxes and their intersection used for computing the intersection points are also shown. There are many general and efficient algorithms available for localizing geometry based on bounding volume hierarchies [8,12,14,19]. However, for exact intersection testing these algorithms typically perform well only on static geometry where the hierarchy can be precomputed. In order to handle dynamic deformable geometry with no precomputation, we use coarse levels for efficient trivial rejection and to obtain reasonable ....

[Article contains additional citation context not shown here]

S. Quinlan, Efficient Distance Computation between Non-Convex Objects. International Conf. on Robotica and Automation, 3324-3329, 1994

....for performing proximity queries between general polyhedra. Specifically, sphere trees, cone trees, axis aligned box trees, oriented box trees, k d trees and octrees, trees based on S bounds, and k dops have been used for fast intersection queries for general polyhedra, as well as polygon soups [13, 20, 11, 21, 1, 14]. For most scenarios, these hierarchies excel at intersection detection but do not do so well when it comes to distance computation. 2.3 Multiresolution Techniques Multiresolution modeling techniques, such as model simplification, have been proposed to extract the shape of the underlying ....

S. Quinlan. Efficient distance computation between non-convex objects. In Proceedings of International Conference on Robotics and Automation, pages 3324--3329, 1994.

....the geometry of A and the obstacles in semi algebraic form, their time complexity makes it impractical for any sufficiently large value of n. On the other hand, reasonably efficient algorithms are available to compute the Euclidean distance between two objects in threedimensional space (e.g. [27]) This leads us to assume that C free is implicitly given by a function, clearance: C R, that maps any configuration to the distance between the robot (placed at that configuration) and the obstacles, if there is no contact between them, and to 0 (or a negative number) otherwise. Thus, ....

S. Quinlan, Efficient Distance Computation Between Non-Convex Objects. Proc. IEEE Intl. Conf. on Robotics and Automation, 3324-3330, San Diego, CA, 1994.

....In the current computer aided manual approach, however, there is no precise way for a user to input access motions that satisfy these constraints. For the first characteristic about model complexity and precise collision checking, we employed the hierarchical spheretree algorithm reported by Quinlan (1994) to reduce the number of calls to the exact collision checking routine for two polygons. The exact collision checking routine is modeled after Gilbert and Johnson (1985) It allows the user to specify desirable accuracy up to double floating point precision. For the second characteristic, we use ....

Quinlan, S., 1994, "Efficient distance computation between non-convex objects," Proc. IEEE Int. Conf. on Rob. and Aut., San Diego, CA.

....by hierarchical spheres Figure 3: Sphere Characterization From measurement, we observed that approximately 50 of the CPU time is spent in detecting when the proxy collides with an object in the graphics display. Collision detection is achieved by an algorithmic approach first described in [8]. The basic idea is to take a polygonal surface and cover each polygon with a small sphere. Then, from this initial set of spheres, they are hierarchically covered. Figure 3 shows the beginnings of covering a face using this method (the actual algorithm was written for three dimensions) At the ....

S. Quinlan, "Efficient Distance Computation between Non-Convex Objects, " International Conference on Robotics and Automation, pp. 3324-3329, 1994.

....geometric models are not convex, and the existence of contaminated data (as called polygon soups ) are also common. Many general algorithms for nonconvex objects have been proposed, and almost all approaches utilize some sort of BV hierarchy. For example, sphere trees were developed in [8][15][16] Although spheres do not bound a polygon as tight as other BV s, sphere trees are easy to implement and efficient at rejection tests. Oriented bounding box (OBB) is another good BV shown to be effective[7] Spherical shell is a BV possessing even better tightness and smaller update cost[11] ....

....to the branch prediction strategy used in prefetching pipelined instructions in modern CPU designs. Its effectiveness will be shown in the next section. 5. EXPERIMENTAL RESULTS 5. 1 Implementation We have implemented a collision detection library based on the sphere tree structure proposed in [15]. The code was written in the C language and runs on most UNIX operating systems. The separation list of the recursion tree was implemented as a linked list of nodes comprised of sphere pairs. The code is also used in a path planner based on artificial potential fields (similar to the planner in ....

S. Quinlan, "Efficient Distance Computation between Non-Convex Objects," Proceedings of International Conference on Robotics and Automation, pp33243329, San Diego, CA, 1994.

....oriented bounding box [GLM96] a k DOP [KHM 98] or a swept sphere volume [LGLM99] These algorithms can only compute all pairs of overlapping triangles and not the intersection region. Algorithms for separation distance computation based on boundingvolume hierarchies have been proposed by [Qui94, JC98, LGLM99] Intersection Region and Penetration Depth: Given two polyhedral models, algorithms for intersection computation and boundary evaluation have been extensively studied in solid modeling. Bouma and Vanecek [BV93] used spatial partitioning approaches to compute the contact region. ....

S. Quinlan. Efficient distance computation between non-convex objects. In Proceedings of International Conference on Robotics and Automation, pages 3324--3329, 1994.

....original object as nodes in the tree. Such a data structure is simple to produce automatically, and lends itself to efficient and elegant recursive algorithms. The disadvantage of this approach is that each level of the hierarchy does not fit the underlying object very tightly. ffl Sphere Trees [13, 25, 26]. The main advantages of using spheres are that they are rotationally invariant, making them very fast to update, and it is very simple to test for distances between them, and test for overlaps. The disadvantage is that spheres do not approximate certain types of objects very efficiently. Hubbard ....

S. Quinlan. Efficient distance computation between non-convex objects. In Edna Straub and Regina Spencer Sipple, editors, Proceedings of the International Conference on Robotics and Automation. Volume 4, pages 3324--3330, Los Alamitos, CA, USA, May 1994. IEEE Computer Society Press.

....in the literature are based upon computing convex decompositions of polyhedra. Given convex decompositions, they combine algorithms for convex polyhedra with algorithms for multiple object pairs. Such algorithms for collision detection and distance computation are presented in [LMC94] and [Qui94], respectively. However, the problem computing the optimal convex decomposition of general polyhedra is NP hard [Cha84] Algorithms of complexity O(n 3 ) are known in the literature [CP90] however they are hard to implement and not much is known about their performance on actual models. The ....

S. Quinlan. Efficient distance computation between nonconvex objects. In Proceedings of International Conference on Robotics and Automation, pages 3324--3329, 1994.

....the root node covers the entire model. It is permissible that child bounding volumes extend beyond the scope of their parent, although for some hierarchies this never occurs, such as with axis aligned bounding boxes and k DOPs. The most commonly used bounding volume types are spheres [Hub93, Qui94] and axis aligned bounding boxes. These shapes are very simple and can be tested for overlap very quickly. More recent work appears to focus on tighter fitting bounding volumes, which usually require fewer bounding volume overlap tests to determine contact between models. However, these more ....

S. Quinlan. Efficient distance computation between non-convex objects. In Proceedings of International Conference on Robotics and Automation, pages 3324--3329, 1994.

....restarting mechanism. z( t t g( Figure 3: Curves of distance extrema embedded into the finger surface model and the CAD model. 2 BACKGROUND The robotics community has considerable literature in the area of finding the minimal distance between a point and model or between model and model [Quinlan, 1994, Gilbert, 1988, Lin, 1994] However, the minimum distance between models is zero during penetration; we desire the penetration depth for haptic force computations. The minimum distance between parametric surfaces f(u, v) and g(s, t) may be described by the following system of equations (f(u, v) ....

Quinlan, S. "Efficient Distance Computation between Non-Convex Objects," IEEE Int. Conference on Robotics and Automation, pp. 3324-3329, 1994.

....for performing proximity queries between general polyhedra. Specifically, sphere trees, cone trees, axis aligned box trees, oriented box trees, k d trees and octrees, trees based on S bounds, and k dops have been used for fast intersection queries for general polyhedra, as well as polygon soups [13, 20, 11, 21, 1, 14]. For most scenarios, these hierarchies excel at intersection detection but do not do so well when it comes to distance computation. 2.3 Multiresolution Techniques Multiresolution modeling techniques, such as model simplification, have been proposed to extract the shape of the underlying ....

S. Quinlan. Efficient distance computation between non-convex objects. In Proceedings of International Conference on Robotics and Automation, pages 3324--3329, 1994.

....used for collision detection between general polygonal models. Typical examples of bounding volumes include axis aligned boxes (cubes are a special case) and spheres, and they are chosen for their fast overlap tests. Other structures include cone trees, k d trees and octrees [43] sphere trees [19,41], trees based on S bounds [6] etc. Binary space partitions (BSP) 39] and extensions to multi space partitions [4] and spatial partitionings based on space time bounds or four dimensional testing [5,9,13,19]have been used. All of these hierarchical methods do very well in performing rejection ....

....contains all the primitives of a model, and children BVs each contain separate partitions of the primitives enclosed by the parent. Leaf node BVs typically contain one primitive. In some variations, one may place several primitivesataleafnode, or use several volumes to contain a single primitive[41]. Two models are compared using a tandem recursive traversal of their BVHs. Each recursive call is made with BVs A and B,say one from each hierarchy,and must determine if A and B overlap. If A and B are not overlapping, the recursion stops at that level. But if A and B are overlapping, the ....

Quinlan, S. (1994). Efficient distance computation between non-convex object, in the Proceedings of IEEE International Conference on Robotics and Automation, pp. 3324-- 3329, 1994.

....robots amid moving obstacles. Our previous experience with PRM planners indicates that the random sampling planning approach scales up well with both complex geometry and many dofs. Geometric complexity essentially increases the cost of collision checking, but hierarchical techniques (e.g. [9, 16]) deal with this issue well. We have successfully applied a randomized path planner to environments with up to 200,000 triangles [11] Another important future direction of research is to integrate the planner with the controller and the sensing modules that detect moving obstacles. The ....

S. Quinlan. Efficient distance computation between nonconvex objects. In Proc. IEEE Int. Conf. on Robotics and Automation, pages 3324--3329, 1994.

No context found.

S. Quinlan. Efficient distance computation between non-convex objects. In Proceedings of IEEE International Conference on Robotics and Automation, volume 4, pages 3324--3329, 1994.

No context found.

S. Quinlan. Efficient distance computation between non-convex objects. In IEEE International Conference on Robotics and Automation, pages 3324--3329, San Diego, CA, May 1994.

No context found.

S. Quinlan, "Efficient distance computation between non-convex objects, " in Proceedings of the 1994.

No context found.

Quinlan, S. Efficient Distance Computation between Non-Convex Objects. In Proc. Int. Conf. on Robotics & Automation, pp.33243329, 1994.

No context found.

S. Quinlan, "Efficient distance computation between nonconvex objects," in IEEE International Conference on Robotics and Automation (ICRA), 1994, pp. 3324--3329.

No context found.

Quinlan, Sean. "Efficient Distance Computation between Non-Convex Objects," IEEE Int. Conference on Robotics and Automation, pp. 3324-3329, 1994.

No context found.

S. Quinlan. Efficient Distance Computation Between Non-Convex Objects. Proc. IEEE Int. Conf. On Robotics and Automation, pp. 3324-3329, 1994.

No context found.

Quinlan S. Efficient distance computation between non-convex objects. Proceedings of the 1994 IEEE International Conference on Robotics and Automation, vol. 4, San Diego, CA, USA, 8-13 May 1994.

No context found.

S. Quinlan. Efficient distance computation between non-convex objects. In Proceedings of IEEE International Conference on Robotics and Automation, volume 4, pages 3324--3329, 1994.

No context found.

S. Quinlan, "Efficient distance computation between non--convex objects," Proceedings of IEEE Int. Conf. on Robotics and Automation, pp. 3324-3329, 1994.

No context found.

S. Quinlan. Efficient distance computation between non--convex objects, Proceedings of IEEE International Conference on Robotics and Automation '94), 3324--3329, 1994. 2

No context found.

S. Quinlan, "Efficient Distance Computation between Non-Convex Objects," Proc. of Intl. Conf. on Robotics and Automation, pp.3324-3329, San Diego, CA, 1994.

No context found.

S. Quinlan. Efficient distance computation between non-convex objects. Proc. IEEE ICRA, 3324--3329, 1994.

No context found.

S. Quinlan. Efficient distance computation between nonconvex objects. Proc. IEEE ICRA, 3324--3329, 1994. 3

No context found.

S. Quinlan. Efficient distance computation between non-convex objects. In Proceedings of International Conference on Robotics and Automation, pages 3324-- 3329, 1994.

First 50 documents  Next 50

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC