Device Nodes

Nodes are the building blocks when populating the Pipeline. Each node provides a specific functionality on the DepthAI, a set of configurable properties and inputs/outputs. After you create a node on a pipeline, you can also configure it as desired and link it to other nodes.A device node is a special type of node, which is run on device SoC (RVC) utilizing the resources of the device. Device nodes are used to offload the host from processing tasks that can be done on the device itself. Alternatively, host nodes can be used to perform computation on the host (PC).On the table of contents (left side of the page) all nodes are listed under the Node entry. You can click on them to find out more.

Inputs and outputs

Each node can have zero, one or multiple inputs and outputs. For example, SystemLogger node has no inputs and 1 output and EdgeDetector has 2 inputs and 1 output (as shown below). Script node can have any number of inputs/ouputs.

React Flow

Each node's output has its pool; block in RAM where it stores messages. Each node's input has its queue for message pointers (doesn't store the message itself, just a pointer to it).

Node input

Node input queue is a queue for Messages. It can be linked with other node's output (that's how you link up nodes). Node inputs are configurable - with input.setBlocking(bool) and input.setQueueSize(num), eg. edgeDetector.inputImage.setQueueSize(10). If the input queue fills up, behavior of the input depends on blocking attribute.Let's say we have linked ColorCamera preview output with NeuralNetwork input input.

If input is set to blocking mode, and input queue fills up, no new messages from ColorCamera will be able to enter the input queue. This means ColorCamera will block and wait with sending its messages until it can push the message to the queue of NeuralNetwork input. If ColorCamera preview is connected to multiple inputs, the same behavior implies, with the messages being pushed sequentially to each input.If blocking is disabled, new messages will push out old messages. This eliminates the risk of pipeline freezing, but can result in dropped messages (eg. ImgFrame).

Node output

Node outputs Messages. Some nodes have a configurable output message pool. Output message pool is a reserved memory region (to reduce memory fragmentation) that holds output messages. After the node creates an output message (for example ImgFrame), it will send it to other nodes as specified when linking the inputs/outputs of the node. Currently, some nodes (VideoEncoder, NeuralNetwork, ImageManip, XLinkIn) can have the pool size configured. The size of the pool specifies how many messages can be created and sent out while other messages are already somewhere in the pipeline.

Pipeline freezing

When all the messages from pool are sent out and none yet returned, that's when the node will block (freeze) and wait until a message is released (not used by any node in the pipeline).

class

dai::Node

#include Node.hppDerived by dai::NodeGroup, dai::ThreadedNode

variable

Id id

variable

std::string alias

function

Node(const Node &)

function

Node & operator=(const Node &)

function

Node(Node &&)

function

Node & operator=(Node &&)

function

Pipeline getParentPipeline()

function

const Pipeline getParentPipeline()

inline function

std::string getAlias()

inline function

void setAlias(std::string alias)

function

const char * getName()

inline function

void start()

inline function

void wait()

inline function

void stop()

function

void stopPipeline()

function

void buildStage1()

function

void buildStage2()

function

void buildStage3()

function

std::vector< Output > getOutputs()

function

std::vector< Input > getInputs()

function

std::vector< Output * > getOutputRefs()

function

std::vector< const Output * > getOutputRefs()

function

std::vector< Input * > getInputRefs()

function

std::vector< const Input * > getInputRefs()

function

std::vector< OutputMap * > getOutputMapRefs()

function

std::vector< InputMap * > getInputMapRefs()

function

Output * getOutputRef(std::string name)

function

Output * getOutputRef(std::string group, std::string name)

function

Input * getInputRef(std::string name)

function

Input * getInputRef(std::string group, std::string name)

function

OutputMap * getOutputMapRef(std::string group)

function

InputMap * getInputMapRef(std::string group)

function

bool isSourceNode()

function

~Node()

function

const AssetManager & getAssetManager()

function

AssetManager & getAssetManager()

function

std::vector< uint8_t > loadResource(dai::Path uri)

function

std::vector< uint8_t > moveResource(dai::Path uri)

inline function

std::shared_ptr< N > create()

function

void add(std::shared_ptr< Node > node)

function

std::vector< std::shared_ptr< Node > > getAllNodes()

function

std::shared_ptr< const Node > getNode(Node::Id id)

function

std::shared_ptr< Node > getNode(Node::Id id)

function

void remove(std::shared_ptr< Node > node)

function

ConnectionMap getConnectionMap()

function

void link(const Node::Output & out, const Node::Input & in)

function

void unlink(const Node::Output & out, const Node::Input & in)

function

void link(std::shared_ptr< Node > in)

function

Node::Output * requestOutput(const Capability & capability, bool onHost)

function

std::vector< std::pair< Input &, std::shared_ptr< Capability > > > getRequiredInputs()

function

bool runOnHost()

inline function

const NodeMap & getNodeMap()

inline function

void buildInternal()

struct

dai::Node::Connection

#include Node.hpp

variable

Id outputId

variable

std::string outputName

variable

std::string outputGroup

variable

Id inputId

variable

std::string inputName

variable

std::string inputGroup

function

Connection(Output out, Input in)

function

Connection(ConnectionInternal c)

function

bool operator==(const Connection & rhs)

struct

dai::Node::ConnectionInternal

#include Node.hpp

variable

std::weak_ptr< Node > outputNode

variable

std::string outputName

variable

std::string outputGroup

variable

std::weak_ptr< Node > inputNode

variable

std::string inputName

variable

std::string inputGroup

variable

Output * out

variable

Input * in

function

ConnectionInternal(Output & out, Input & in)

function

bool operator==(const ConnectionInternal & rhs)

struct

dai::Node::ConnectionInternal::Hash

function

size_t operator()(const dai::Node::ConnectionInternal & obj)

struct

dai::Node::DatatypeHierarchy

variable

DatatypeEnum datatype

variable

bool descendants

inline function

DatatypeHierarchy(DatatypeEnum d, bool c)

class

dai::Node::Input

variable

std::vector< DatatypeHierarchy > possibleDatatypes

inline explicit function

Input(Node & par, InputDescription desc, bool ref)

inline function

const Node & getParent()

Get the parent node

inline function

Node & getParent()

Get the parent node

inline function

Type getType()

Get type

function

std::string toString()

Input to string representation

function

void setWaitForMessage(bool waitForMessage)

Overrides default wait for message behavior. Applicable for nodes with multiple inputs. Specifies behavior whether to wait for this input when a Node processes certain data or not.

Parameters

waitForMessage: Whether to wait for message to arrive to this input or not

function

bool getWaitForMessage()

Get behavior whether to wait for this input when a Node processes certain data or not

Returns

Whether to wait for message to arrive to this input or not

function

std::vector< DatatypeHierarchy > getPossibleDatatypes()

Get possible datatypes that can be received

function

void setPossibleDatatypes(std::vector< DatatypeHierarchy > types)

Set possible datatypes that can be received

function

void setReusePreviousMessage(bool reusePreviousMessage)

Equivalent to setWaitForMessage but with inverted logic.

function

bool getReusePreviousMessage()

Equivalent to getWaitForMessage but with inverted logic.

function

void setGroup(std::string group)

Set group name for this input

function

std::string getGroup()

Get group name for this input

function

bool isConnected()

Check if this input is connected

function

std::shared_ptr< InputQueue > createInputQueue(unsigned int maxSize, bool blocking)

enum

Type

enumerator

SReceiver

enumerator

MReceiver

struct

dai::Node::InputDescription

variable

std::string name

variable

std::string group

variable

bool blocking

variable

int queueSize

variable

std::vector< DatatypeHierarchy > types DEFAULT_TYPES

variable

bool waitForMessage

class

dai::Node::InputMap

#include Node.hpp

variable

std::string name

function

InputMap(Node & parent, InputDescription defaultInput)

function

InputMap(Node & parent, std::string name, InputDescription defaultInput)

function

Input & operator[](const std::string & key)

function

Input & operator[](std::pair< std::string, std::string > groupKey)

function

bool has(const std::string & key)

class

dai::Node::Output

function

~Output()

inline function

Output(Node & par, OutputDescription desc, bool ref)

inline function

Node & getParent()

inline function

const Node & getParent()

function

std::string toString()

inline function

std::string getName()

Get name of the output

inline function

std::string getGroup()

Get group of the output

inline function

void setGroup(std::string group)

Set group name for this output

inline function

void setName(std::string name)

Set name for this output

inline function

Type getType()

Get type of the output

function

std::vector< DatatypeHierarchy > getPossibleDatatypes()

Get possible datatypes that can be sent

function

void setPossibleDatatypes(std::vector< DatatypeHierarchy > types)

Set possible datatypes that can be sent

function

bool isSamePipeline(const Input & in)

Check if this output and given input are on the same pipeline.

Parameters

canConnect for checking if connection is possible

Returns

True if output and input are on the same pipeline

function

bool canConnect(const Input & in)

Check if connection is possible

Parameters

in: Input to connect to

Returns

True if connection is possible, false otherwise

function

std::vector< ConnectionInternal > getConnections()

Retrieve all connections from this output

Returns

Vector of connections

inline function

std::vector< QueueConnection > getQueueConnections()

Retrieve all queue connections from this output

Returns

Vector of queue connections

function

std::shared_ptr< dai::MessageQueue > createOutputQueue(unsigned int maxSize, bool blocking)

function

void link(Input & in)

Link current output to input.Throws an error if this output cannot be linked to given input, or if they are already linked

Parameters

in: Input to link to

function

void link(std::shared_ptr< Node > in)

function

void unlink(Input & in)

Unlink a previously linked connectionThrows an error if not linked.

Parameters

in: Input from which to unlink from

function

void send(const std::shared_ptr< ADatatype > & msg)

Sends a Message to all connected inputs

Parameters

msg: Message to send to all connected inputs

function

bool trySend(const std::shared_ptr< ADatatype > & msg)

Try sending a message to all connected inputs

Parameters

msg: Message to send to all connected inputs

Returns

True if ALL connected inputs got the message, false otherwise

struct

dai::Node::Output::QueueConnection

variable

Output * output

variable

std::shared_ptr< MessageQueue > queue

inline function

bool operator==(const QueueConnection & rhs)

enum

Type

enumerator

MSender

enumerator

SSender

struct

dai::Node::OutputDescription

variable

std::string name

variable

std::string group

variable

std::vector< DatatypeHierarchy > types DEFAULT_TYPES

class

dai::Node::OutputMap

#include Node.hpp

variable

std::string name

function

OutputMap(Node & parent, std::string name, OutputDescription defaultOutput, bool ref)

function

OutputMap(Node & parent, OutputDescription defaultOutput, bool ref)

function

Output & operator[](const std::string & key)

function

Output & operator[](std::pair< std::string, std::string > groupKey)

struct

dai::Node::PairHash

inline function

std::size_t operator()(const std::pair< T1, T2 > & pair)

enum

ON THIS PAGE

Device Nodes

Inputs and outputs

Node input

Node output

dai::Node

dai::Node::Connection

dai::Node::ConnectionInternal

dai::Node::ConnectionInternal::Hash

dai::Node::DatatypeHierarchy

dai::Node::Input

Type

dai::Node::InputDescription

dai::Node::InputMap

dai::Node::Output

dai::Node::Output::QueueConnection

Type

dai::Node::OutputDescription

dai::Node::OutputMap

dai::Node::PairHash

std::int64_t Id