* * \brief List of points * * Stores a list of *n* points (3 x n matrix) together with their surface * normals * (3 x n matrix). Also keeps infor

* * \brief Check if a grasp is antipodal. * * This class checks if a grasp candidate satisfies the antipodal condition. * */

*\brief 2-D bounding box of hand closing region with respect to hand frame */

* * \brief Classify grasp candidates as viable grasps or not with Caffe * * Classifies grasps as viable or not using a convolutional neural network

* * \brief Generate grasp candidates. * * This class generates grasp candidates by searching for feasible robot hand * placements in a point cloud

* * \brief Abstract base class for classifier that classifies grasp candidates as * viable grasps or not. * */

* * \brief Multi-view point cloud * * Stores and processes a point cloud that has been observed from one or * multiple camera view points. The raw

* * \brief Group grasp candidates in clusters * * This class searches for clusters of grasps. Grasps in the same cluster are * geometrically simil

* * \brief Configuration file * * Reads parameters from a configuration file (`*.cfg`). The configuration file * is a key-value storage. * */

* * \brief Convolutional layer. * * A convolutional layer for a neural network for the `EigenClassifier`. * */

* * \brief Dense (fully connected) layer. * * A dense (fully connected) layer for a neural network for the * `EigenClassifier`. * */

* * \brief Classify grasp candidates as viable grasps or not with Eigen * * Classifies grasps as viable or not using a custom neural network framew

* * \brief Calculate collision-free finger placements. * * This class calculates collision-free finger placements. The parameters are * the outer

* * \brief Estimate local reference frames. * * This class estimates local reference frames (LRFs) for point neighborhoods. * */

* * \brief Detect grasp poses in point clouds. * * This class detects grasp poses in a point clouds by first creating a large * set of grasp candi

* * \brief Grasp represented as a robot hand pose * * This class represents a grasp candidate by the position and orientation of * the robot hand

* * \brief Store robot hand geometry * * This class stores parameters which define the geometry of the robot hand. * This geometry is used to calc

* * \brief Search for grasp candidates. * * This class searches for grasp candidates in a point cloud by first * calculating a local reference fra

* * \brief Calculate a set of grasp candidates. * * This class calculate sets of grasp candidates. The grasp candidates in the * same set share th

* * \brief Calculate 12-channels grasp image. * * The 12 channels contain height maps and surface normals of the points * contained inside the rob

* * \brief Calculate 15-channels grasp image. * * The 15 channels contain height maps and surface normals of the points * contained inside the rob

* * \brief Calculate binary grasp image. * * The binary image represents the shape of what is contained inside the robot * hand's closing region.

* * \brief Calculate 3-channels grasp image. * * The 3 channels contain the surface normals of the points contained inside * the robot hand's clos

* * \brief Create grasp images for classification. * * Creates images for the input layer of a convolutional neural network. Each * image represen

* * \brief Store grasp image geometry. * * Stores the parameters used to calculate a grasp image. * * Each grasp image is based on a bounding box

* * \brief Abstract base class for calculating grasp images/descriptors. * * Also offers methods to calculate various images. * */

*\brief Label information */

* * \brief Abstract base class for neural network layers in the custom framework. * */

* * \brief Local reference frame. * * This class estimates the local reference frame (LRF) for a point * neighborhood. The coordinate axes of the

* * \brief Visualization utilities * * Provides visualization methods that use the PCL Visualizer. Allows to * visualize samples, surface normals,

* * \brief Grasp pose detection with the Cross Entropy Method. * * This class uses the Cross Entropy Method to focus the grasp candidate * generat