From the Publisher: 
The accessible presentation of this book gives both a general view of the entire computer vision enterprise and also offers sufficient detail to be able to build useful applications. Users learn techniques that have proven to be useful by first-hand experience and a wide range of mathematical methods. A CD-ROM with every copy of the text contains source code for programming practice, color images, and illustrative movies. Comprehensive and up-to-date, this book includes essential topics that either reflect practical significance or are of theoretical importance. Topics are discussed in substantial and increasing depth. Application surveys describe numerous important application areas such as image based rendering and digital libraries. Many important algorithms broken down and illustrated in pseudo code. Appropriate for use by engineers as a comprehensive reference to the computer vision enterprise.

Computer vision : a modern approach = 计算机视觉 : 一种现代的方法

We present a new approach to motion planning using a stochastic trajectory optimization framework. The approach relies on generating noisy trajectories to explore the space around an initial (possibly infeasible) trajectory, which are then combined to produced an updated trajectory with lower cost. A cost function based on a combination of obstacle and smoothness cost is optimized in each iteration. No gradient information is required for the particular optimization algorithm that we use and so general costs for which derivatives may not be available (e.g. costs corresponding to constraints and motor torques) can be included in the cost function. We demonstrate the approach both in simulation and on a mobile manipulation system for unconstrained and constrained tasks. We experimentally show that the stochastic nature of STOMP allows it to overcome local minima that gradient-based methods like CHOMP can get stuck in.

STOMP: Stochastic trajectory optimization for motion planning

We present a new optimization-based approach for robotic motion planning among obstacles. Like CHOMP (Covariant Hamiltonian Optimization for Motion Planning), our algorithm can be used to find collision-free trajectories from naA¯ve, straight-line initializations that might be in collision. At the core of our approach are (a) a sequential convex optimization procedure, which penalizes collisions with a hinge loss and increases the penalty coefficients in an outer loop as necessary, and (b) an efficient formulation of the no-collisions constraint that directly considers continuous-time safety Our algorithm is implemented in a software package called TrajOpt.

We report results from a series of experiments comparing TrajOpt with CHOMP and randomized planners from OMPL, with regard to planning time and path quality. We consider motion planning for 7 DOF robot arms, 18 DOF full-body robots, statically stable walking motion for the 34 DOF Atlas humanoid robot, and physical experiments with the 18 DOF PR2. We also apply TrajOpt to plan curvature-constrained steerable needle trajectories in the SE(3) configuration space and multiple non-intersecting curved channels within 3D-printed implants for intracavitary brachytherapy. Details, videos, and source code are freely available at: http://rll.berkeley.edu/trajopt/ijrr.

/pdf/motion-planning-with-sequential-convex-optimization-and-4hyn1dc3bw.pdf

Motion planning with sequential convex optimization and convex collision checking

In one embodiment of the invention, a method for a robotic system is disclosed to track one or more robotic instruments. The method includes generating kinematics information for the robotic instrument within a field of view of a camera; capturing image information in the field of view of the camera; and adaptively fusing the kinematics information and the image information together to determine pose information of the robotic instrument. Additionally disclosed is a robotic medical system with a tool tracking sub-system. The tool tracking sub-system receives raw kinematics information and video image information of the robotic instrument to generate corrected kinematics information for the robotic instrument by adaptively fusing the raw kinematics information and the video image information together.

Methods and systems for robotic instrument tool tracking with adaptive fusion of kinematics information and image information

Society is becoming more automated with robots beginning to perform most tasks in factories and starting to help out in home and office environments. One of the most important functions of robots is the ability to manipulate objects in their environment. Because the space of possible robot designs, sensor modalities, and target tasks is huge, researchers end up having to manually create many models, databases, and programs for their specific task, an effort that is repeated whenever the task changes. Given a specification for a robot and a task, the presented framework automatically constructs the necessary databases and programs required for the robot to reliably execute manipulation tasks. It includes contributions in three major components that are critical for manipulation tasks. 
The first is a geometric-based planning system that analyzes all necessary modalities of manipulation planning and offers efficient algorithms to formulate and solve them. This allows identification of the necessary information needed from the task and robot specifications. Using this set of analyses, we build a planning knowledge-base that allows informative geometric reasoning about the structure of the scene and the robot's goals. We show how to efficiently generate and query the information for planners. 
The second is a set of efficient algorithms considering the visibility of objects in cameras when choosing manipulation goals. We show results with several robot platforms using grippers cameras to boost accuracy of the detected objects and to reliably complete the tasks. Furthermore, we use the presented planning and visibility infrastructure to develop a completely automated extrinsic camera calibration method and a method for detecting insufficient calibration data. 
The third is a vision-centric database that can analyze a rigid object's surface for stable and discriminable features to be used in pose extraction programs. Furthermore, we show work towards a new voting-based object pose extraction algorithm that does not rely on 2D/3D feature correspondences and thus reduces the early-commitment problem plaguing the generality of traditional vision-based pose extraction algorithms. 
In order to reinforce our theoric contributions with a solid implementation basis, we discuss the open-source planning environment OpenRAVE, which began and evolved as a result of the work done in this thesis. We present an analysis of its architecture and provide insight for successful robotics software environments.

Automated construction of robotic manipulation programs

Simulation of the musculoskeletal system has important applications in biomechanics, biomedical engineering, surgery simulation, and computer graphics. The accuracy of the muscle, bone, and tendon geometry as well as the accuracy of muscle and tendon dynamic deformation are of paramount importance in all these applications. We present a framework for extracting and simulating high resolution musculoskeletal geometry from the segmented visible human data set. We simulate 30 contact/collision coupled muscles in the upper limb and describe a computationally tractable implementation using an embedded mesh framework. Muscle geometry is embedded in a nonmanifold, connectivity preserving simulation mesh molded out of a lower resolution BCC lattice containing identical, well-shaped elements, leading to a relaxed time step restriction for stability and, thus, reduced computational cost. The muscles are endowed with a transversely isotropic, quasiincompressible constitutive model that incorporates muscle fiber fields as well as passive and active components. The simulation takes advantage of a new robust finite element technique that handles both degenerate and inverted tetrahedra.

/pdf/creating-and-simulating-skeletal-muscle-from-the-visible-5nk7kxd7ef.pdf

Creating and simulating skeletal muscle from the visible human data set

Robots that perform complex manipulation tasks must be able to generate strategies that make and break contact with the object. This requires reasoning in a motion space with a particular multi-modal structure, in which the state contains both a discrete mode (the contact state) and a continuous configuration (the robot and object poses). In this paper we address multi-modal motion planning in the common setting where the state is high-dimensional, and there are a continuous infinity of modes. We present a highly general algorithm, Random-MMP, that repeatedly attempts mode switches sampled at random. A major theoretical result is that Random-MMP is formally reliable and scalable, and its running time depends on certain properties of the multi-modal structure of the problem that are not explicitly dependent on dimensionality. We apply the planner to a manipulation task on the Honda humanoid robot, where the robot is asked to push an object to a desired location on a cluttered table, and the robot is restricted to switch between walking, reaching, and pushing modes. Experiments in simulation and on the real robot demonstrate that Random-MMP solves problem instances that require several carefully chosen pushes in minutes on a PC.

Randomized multi-modal motion planning for a humanoid robot manipulation task

The purpose of this study was to visualize and document the architecture of the human soleus muscle throughout its entire volume. The architecture was visualized by creating a threedimensional (3D) manipulatable computer model of an entire cadaveric soleus, in situ, using B-spline solid to display muscle fiber bundles that had been serially dissected, pinned, and digitized. A database of fiber bundle length and angle of pennation throughout the marginal, posterior, and anterior soleus was compiled. The computer model allowed documentation of the architectural parameters in 3D space, with the angle of pennation being measured relative to the tangent plane of the point of attachment of a fiber bundle. Before this study, the only architectural parameters that have been recorded have been 2D. Three-dimensional reconstruction is an exciting innovation because it makes feasible the creation of an architectural database and allows visualization of each fiber bundle in situ from any perspective. It was concluded that the architecture is non-uniform throughout the volume of soleus. Detailed architectural studies may lead to the development of muscle models that can more accurately predict interaction between muscle parts, force generation, and the effect of pathologic states on muscle function. Clin. Anat. 16:285‐293, 2003. © 2003 Wiley-Liss, Inc.

Documentation and three-dimensional modelling of human soleus muscle architecture.

This paper considers a shortcutting heuristic to smooth jerky trajectories for many-DOF robot manipulators subject to collision constraints, velocity bounds, and acceleration bounds. The heuristic repeatedly picks two points on the trajectory and attempts to replace the intermediate trajectory with a shorter, collision-free segment. Here, we construct segments that interpolate between endpoints with specified velocity in a time-optimal fashion, while respecting velocity and acceleration bounds. These trajectory segments consist of parabolic and straight-line curves, and can be computed in closed form. Experiments on reaching tasks in cluttered human environments demonstrate that the technique can generate smooth, collision-free, and natural-looking motion in seconds for a PUMA manipulator and the Honda ASIMO robot.

/pdf/fast-smoothing-of-manipulator-trajectories-using-optimal-fxktl6q0dd.pdf

Fast smoothing of manipulator trajectories using optimal bounded-acceleration shortcuts

We present a model that is capable of synchronizing expressive gestures with speech. The model, implemented on a Honda humanoid robot, can generate a full range of gesture types, such as emblems, iconic and metaphoric gestures, deictic pointing and beat gestures. Arbitrary input text is analyzed with a part-of-speech tagger and a text-to-speech engine for timing information of spoken words. In addition, style tags can be optionally added to specify the level of excitement or topic changes. The text, combined with any tags, is then processed by several grammars, one for each gesture type to produce several candidate gestures for each word of the text. The model then selects probabilistically amongst the gesture types based on the desired degree of expressivity. Once a gesture type is selected, it coincides with a particular gesture template, consisting of trajectory curves that define the gesture. Speech timing patterns and style parameters are used to modulate the shape of the curve before it sent to the whole body control system on the robot. Evaluation of the model's parameters were performed, demonstrating the ability of observers to differentiate varying levels of expressiveness, excitement and speech synchronization. Modification of gesture speed for trajectory tracking found that positive associations like happiness and excitement accompanied faster speeds, with negative associations like sadness or tiredness occurred at slower speeds.

/pdf/synchronized-gesture-and-speech-production-for-humanoid-2mjki91df5.pdf

Victor Ng-Thow-Hing

Papers

Creating and simulating skeletal muscle from the visible human data set

Randomized multi-modal motion planning for a humanoid robot manipulation task

Documentation and three-dimensional modelling of human soleus muscle architecture.

Fast smoothing of manipulator trajectories using optimal bounded-acceleration shortcuts

Synchronized gesture and speech production for humanoid robots