Module API

A DV module is processing unit that takes some streams as inputs, does operations on them, and outputs streams as outputs. Typically, a module takes on polarity event data, does computer vision algorithms on the data and outputs the results. Input and outputs can have arbitrary types.

Module Basics

Every DV module inherits from the dv::ModuleBase class. In order to register a module class (that inherits from dv::ModuleBase) as a DV module, the following macro has to be called at the end of the file where the class is defined:

registerModuleClass(RefractoryPeriodFilter)

Fields

Inheriting from the this class gives access to the following fields

Field	Type	Purpose
`config`	dv::RuntimeConfig	Access config values from config options.
`log`	dv::Logger	A configured logging instance to log values to DV
`inputs`	dv::RuntimeInputs	Access to the inputs of the module to get data
`outputs`	dv::RuntimeOutputs	Access to the outputs of the module to send data

Static Methods

Every module has to provide some static methods to work properly with DV. The static methods must have the exact footprint as described in the list below. You have to provide these methods.

static const char *initDescription()

Return a description of what the module does This function is required

static void initInputs(dv::InputDefinitionList &in)

Define the inputs to the module

static void initOutputs(dv::OutputDefinitionList &out)

Define the outputs to the module

static void initConfigOptions(dv::RuntimeConfig &config)

Define configuration options for the module

Overridable Methods

dv::BaseModule exhibits two overridable functions to implement the functionality of the module.

void run()

Gets called periodically or as new data arrives. Should do the processing. This function is required

void configUpdate()

Gets called whenever the user changes the configuration. Used to read new configuration values and store them in instance data members.