| ▼NROSEE | |
| ▼NTestUtils | |
| CProcess | |
| ▼NUtils | |
| CDifferentKeysException | |
| CSecondOrderFilter | |
| CTimer | |
| CAction | The pure virtual class representing an Action. It has members that are in common to all derived class |
| CActionComposed | A ActionComposed, which is formed by one or more Primitives (or even other composed). It is useful for example to create an action that grasp only with bending the tips (e.g. to take a dish from above) If the ActionComposed has the boolean value independent to true, it means that include indipendent sub-actions, so, each joint is used by at maximum by ONLY ONE of the sub-action inside. In this case the jointsInvolvedCount will contain only 0 or 1 values. If the ActionComposed is not independent, each joint position is calculated as the mean of all the joint position of the contained sub-actions that uses that joint. So each mean can include different primitives, so we used the jointsInvolvedCount vector to store the number of sub action that use each joint |
| CActionGeneric | Class to handle a generic, simple action. Differently from other class, this is easily creable manually, for ex giving a name and a JointPos map to the costructor. It contains essential infos about the action, and override the necessary pure virtual functions of base class Action |
| ▼CActionMultiplePinchTight | Class to describe the action of "pinching" with more than 2 tips (with 2 tips there is the ActionPinchTight and ActionPinchLoose The number of the finger used is fixed when the object is costructed, and it is stored in the father member nFingersInvolved |
| CdepthComp | Struct to put in order the actionStates. with "<" we put as best the position that has less sumDepth |
| CActionPinchGeneric | A base virtual class for the PinchTight and PinchLoose classes. It includes member and method that are in common between the two type of pinches. It derives the more generic ActionPrimitive |
| ▼CActionPinchLoose | The action of pinch with two tips. The two tips must not collide ever (otherwise we have a TightPinch). They only need to move towards each other moving the relative joints. This PinchLoose is created because also if the tips do not collide (i.e. there is not a ActionPinchTight) we can have anyway a pinch at least to take object of a certain minimum size. All the non involved fingers are set in the default state. A pinchLoose is defined by: |
| CdistComp | Struct to put in order the actionStates set. The first elements are the ones with lesser distance |
| ▼CActionPinchTight | The action of pinch with two tips. The two tips must collide for some hand configuration to mark this configuration as a pinchTight. All the non involved fingers are set in the default state. A pinchTight is defined by: |
| CdepthComp | Struct to put in order the actionStates. The first elements are the ones with greater depth , even if is almost impossible, two different contact with same depth will be considered equal with this definition of depthComp. Theoretically they are equal only if the joint status are equal (of only joints that act for the collision). In fact, we should have the possibility to have two contact with the same depth (if joint statuses are different), they will be equally good |
| CActionPrimitive | Virtual class, Base of all the primitive actions. It has some implemented functions that a derived class can use, if you don't want to override to make them more action-specific. All the primitives are defined by: |
| CActionSingleJointMultipleTips | Primitive which indicate a motion of n fingers moving ONLY ONE joint. For example, this primitive is necessary for the hands that have one joint that close all the fingers to do a grasp. But it can also useful to detect other multiple finger motions (like a "spread finger") |
| CActionTimed | An action composed by other ones that must be executed one after other with some wait time (also 0) in between. E.G. 0.000000second --—> Grasp --—> 0.1 + 0.1 second --—> OpenLid --—> 0.5second -— and so on This class contains all the joint position of each action (actionsJointPosMap ) and some time margins which indicates wait time before and after the action (actionsTimeMarginsMap). Each action inside is identified by its name, so no two action with same name can exist (see insertAction ) After create the ActionTimed object, we can add actions with insertAction(). As all other Action classes, it implements also functions to emit and parse in a yaml file |
| CActionTrig | The action of moving some joints (see later) of a single finger in a full clousure position towards the palm. The action is unique (joints involved in a certain position: the bound) so maxStoredActionStates == 1 always Described by: |
| CDummyHal | Class representing an end-effector |
| CEEHal | Class representing an end-effector |
| CEEInterface | Class representing and End-Effector |
| CFindActions | Class to check which fingertips collide (for the pinch action at the moment) |
| CMapActionHandler | |
| CParser | Class responsible for parsing the YAML file providing the URDF, SRDF, the EE-HAL library implementation and the EEInterface |
| CParserMoveIt | Class to parse urdf and srdf with moveit classes and to give information about the model parsed |
| CRosActionServer | |
| CRosServiceHandler | |
| CUniversalRosEndEffectorExecutor | Class representing the ROS executor implementing the unviversal ROS EE concept |
| CXBot2Hal | Concrete class which communicate directly with ROS topics |
| CYamlWorker | |