Nodes

Nodes are the building blocks of Video Vision business logic. To create a functioning customer solution, the nodes in the flow builder need to be correctly placed, linked, and configured.

Each node has a specific purpose and use. You will find each node’s functionality below.

Inputs

Input nodes serve as data entry points of your flow. They can represent input sources like cameras and relays or custom inputs, such as text or a number.

Boolean Input

Emits one of two possible values (i.e., True or False) into the flow from an external connection (e.g., an ADAM-6060, or from a different flow).

Current Date & Time

Outputs the current date and time.

Custom Image

Outputs a solid color image with a custom color and resolution.

Image Stream

Outputs individual images from a camera into the flow for further processing.

Number

Outputs a custom numeric value based on the node configuration.

Number Input

Returns a number from an external connection or another flow.

Polygon

Outputs a custom polygon based on the configuration.

Schedule

Represents the timetable specified in the Scheduler section of the Video Vision application.

Switch

Outputs a true or false value based on the node configuration.

Text

Outputs custom text based on the node configuration.

Image processors

Image processors modify or transform images in order to make them more suitable for a given solution.

Adjust Polygons To Original Image

Adjusts polygon coordinates detected in a section of an image so they can be correctly displayed on the region of the original image. It handles rescaling and shifting to ensure that the polygon’s position reflects the entire image rather than just the subregion.

This functionality is essential when rescaling images for purposes such as neural network training while needing to overlay the detected polygons on the original image.

Below is a visual representation of the process:

Crop Image

Crops the image on each side and then outputs it modified.

Merge Images

The Merge Images node arranges multiple images into a single image output. The layout of the final image depends on the node configuration.

Overlay Heatmap

The Overlay Heatmap node overlays a heatmap onto an image, allowing you to visualize data directly within the image context. The node includes options to adjust the heatmap’s opacity and select from several predefined color gradients.

Color Gradients

To optimize how heatmaps are displayed on the video vision dashboard, we have selected several industry standard heatmaps.

Polygon Crop

Crops the image to the area of a polygon.

Region of Interest

The Region of Interest node allows you to filter or identify polygons based on their presence in specific areas of the image. For example, suppose a portion of the camera image displays a busy area with people. In that case, it might interfere with the detection accuracy or cause false positives, so the node would be used to detect only the area of production that doesn’t have people in it.

Resize image

The Resize Image node changes the image size to the configured resolution. This can be used to resize the image for the dashboard, to match the resolutions accepted by the AI model, or for reporting purposes.

Rotate image

Rotates the input image clockwise based on the configuration.

Undistort Image

Some cameras introduce distortion to images, which causes some straight lines to appear curved. The Undistort Image node is used to fix this distortion.

For more information about how undistorting works, see this article by OpenCV.

Models

AI Model

The AI Model node represents the machine-learning model downloaded in the Models section of the builder. It receives an image input and processes it using the ML model.

Dynamic AI Model

The Dynamic AI Model node allows for runtime switching between machine-learning models, enabling flexibility in processing different types of image inputs without restarting the deployment.

Unlike the standard AI Model node, which uses a single preloaded model, the Dynamic AI Model node can unload the current model and load a new one as needed. This functionality is especially useful for workflows requiring multiple models to handle varied tasks within the same deployment, as well as situations where the GPU memory cannot support multiple models simultaneously.

Model Processors

Model processor nodes handle the data provided by the machine-learning model.

Annotate Polygons

The Annotate Polygons node highlights all input polygons of objects received from the previous node with a specific color and thickness.

The Annotate Polygons node is used to distinguish between different categories of identified items.

For instance, if an AI model is designed to detect goods with and without defects, you would use two Annotate Polygons nodes: one to color the polygons green and the other to color the polygons red. For customer solutions that need to detect just one object type, you would need only one Annotate Polygons node.

Count Polygons

The Count Polygons node counts the number of detected objects in an image. This node is mainly used to output the total number of polygons to the dashboard via the Number Output node.

Draw Polygons

The Draw Polygons node creates a visible overlay of annotated polygons on the image.

Dynamic Selector

The Dynamic Selector node converts input values to new output values according to predefined configurations. It supports different value types, making it useful for tasks such as mapping specific identifiers to numerical data, or converting between types such as strings and booleans for further processing in production workflows.

Extract Attributes

Extracts a desired attribute from the AI Model node for further processing.

 

Merge Polygons

The Merge Polygons node merges multiple polygon lists into one.

Resize Polygons

Resizes polygons by adjusting their vertical and horizontal dimensions. The polygons are resized from the center.

Track Objects

The Track Objects node is used in customer use cases where items need to be tracked while moving (e.g., on a conveyor belt).

Filters

Boolean Trigger

The Boolean trigger node only passes the input through if a condition is True.

This node is used, e.g., if you only want to process the image when the detected object is present in the image or when a button is pressed

Condition

The condition node checks if a condition is true or false and then forwards the input if it is true.

Consecutive Trigger

The Consecutive Trigger node waits until enough positive (True) inputs are received. Once the number is reached, all positive inputs are passed through. Negative inputs reset the process.

Construct a List

The Construct a List node generates a list by combining multiple inputs, where each input becomes an item in the list.

Event Trigger

Only returns a positive value when the input value changes from negative to positive. Subsequent positive inputs are returned as negative.

Heatmap Threshold
The Heatmap Threshold node sets a threshold to highlight areas of an image as a heatmap. Areas meeting the threshold are shown in white, while those exceeding it are shown in black.

Interval Filter

Sets a time delay between positive inputs.

When a positive (True) input is passed through, all subsequent inputs are returned as negative until the timer ends.

Isolate Polygon

Separates a single polygon from multiple polygons based on the attributes selected in the configuration.

Polygon Threshold

The Polygon Threshold node filters out polygons based on the configuration.

Logic

AND

The AND node is a Logic gate. A logic gate takes input signals and produces an output based on rules specific to the gate.

For AND, if all inputs are true, the output is true (signal is passed). Otherwise, the output is false (signal is blocked).

Boolean to Text

The Boolean to Text node transforms boolean values (true/false) to a text output of your choice.

Usage example: If you want the dashboard to display “OK” or “Faulty” for each item that passes through the camera, you enter “Faulty” in the Negative value field and “OK” in the Positive value field. Then, you connect a node with a ‘bool’ value type output (e.g., Compare Value) to the input of Boolean to Text, and you connect the output of Boolean to Text to a Text Output node.

 

Compare value

The Compare Value node compares inputs via mathematical operators and outputs a True or False value based on the result.

For instance, if you configure the node to “>= 15”, it will output a false value until the input signal is 15 or higher, at which point it will output a true value.

Division

The Division node divides an input number (numerator) by a second input number (denominator), then outputs the result.

NAND

The NAND node is a Logic gate. A logic gate takes input signals and produces an output based on rules specific to the gate.

NAND is an AND gate followed by a negation (the signal is inverted). For NAND, if any inputs are true, the output is false (signal is blocked). Otherwise, the output is true (signal is passed).

NOR

The NOR node is a Logic gate. A logic gate takes input signals and produces an output based on rules specific to the gate.

NOR is an OR gate followed by a negation (the signal is inverted). For NOR, if any of the inputs are true, the output is false (signal is blocked). Otherwise, the output is true (signal is passed).

NOT

The NOT node is a Logic gate. A logic gate takes input signals and produces an output based on rules specific to the gate.

For NOT, any input is inverted. If the input is true, output is false and vice versa.

OR

The OR node is a Logic gate. A logic gate takes input signals and produces an output based on rules specific to the gate.

For OR, if any of the inputs are true, the output is true (signal is passed). Otherwise, the output is false (signal is blocked).

Sum

The Sum node adds up all the numerical inputs and then outputs them as a sum. The node doesn’t store any previous values; all new inputs are added up once.

XOR

The XOR node is a Logic gate. A logic gate takes input signals and produces an output based on rules specific to the gate.

For XOR, if exactly one of the inputs is true, the output is true (signal is passed). Otherwise, the output is false (signal is blocked).

Aggregators

Accumulator

The Accumulator node adds up and stores the numeric value received in the ‘Input’ connection.

The stored value can be reset to zero upon receiving any signal in the ‘reset’ connection.

Counter

The Counter node counts the number of input signals received. This can be used for counting the instances of receiving, e.g., defects, images, or polygons. Every new input increments the current value stored in the node by 1.

Get Shared Memory and Set Shared Memory

The ‘Get Shared Memory’ and ‘Set Shared Memory’ nodes work together to temporarily store and retrieve data.

The ‘Set Shared Memory’ node saves input data using a defined ‘key,' allowing the 'Get Shared Memory’ node to later access this value by referencing the same key. This connection enables data continuity and data transfer between nodes and flows.

Set Shared Memory I/O

Get Shared Memory I/O

Miscellaneous

This category contains various nodes without a specific category.

Barcode Reader

The Barcode Reader node detects barcodes on the image input and outputs values detected in the barcode.

Change Flow

The Change Flow node is used to switch to a different, preconfigured state (represented by a flow). The change to the new flow is triggered by either a bool input with a True value or on all other inputs of any value type.

Hold

Temporarily delays the input for a set amount of time.

Table statistics

Structures data sets into rows and columns.

Timer

The timer node triggers an event after a set period of time.

This node can be used for, e.g., delaying the processing of the image to give lights enough time to activate.

Outputs

All output nodes serve as endpoints of the flow.

These nodes can be used to:

• display information on the customer dashboard

• transfer data to another flow

Boolean Output

The Boolean Output node receives a bool value from the flow to make it displayable on the customer dashboard.

Capture Flow

The Capture Flow node is used to trigger data collection for Reporting.

When this node receives a True boolean signal, it records data from all Output nodes in the flow, provided those nodes are marked for reporting in the Deployment settings. If the received signal is False or absent, none of the selected outputs are stored.

For example, you can collect flow data only when specific conditions are met, such as detecting a defect or exceeding a certain value threshold.

Date & Time

The Date & Time output node receives date and time from the flow to make it displayable on the customer dashboard.

Image

The Image output node receives an image from the flow to make it displayable on the customer dashboard.

Number

The Number output node receives a number from the flow to make it displayable on the customer dashboard.

Point List

The Point List output node is used to output a point list to a different flow.

Polygon

The Polygon output node is used to output a point list to a different flow.

Table

The Table output node receives collected statistics in a table grid to make it displayable on the customer dashboard.

Text

The Text output node receives text from the flow to make it displayable on the customer dashboard.