Introduction

This guide discusses the Multi-Unit alignment functionality available on the Irisys Vector units, and how to configure it. 

A full understanding of the settings below is essential to ensure the high accuracy that the Vector 4D devices are capable of.

Note: for correct wide tracker functionality always ensure devices are updated to latest firmware before configuring. See also separate section physical connection of devices on a multi-unit network.

In order to access these settings, you must first make an initial connection to the Vector unit; see separate guide: Irisys Vector 4D - Making a Setup Connection.

When devices are installed and configured to work together it is important to ensure they are aligned correctly so that tracking across the combined field of view of all the devices works correctly.

Essentially devices need to be 'told' where they are in relation to the neighbours so that people who walk between the individual field of views of each device will be correctly assigned to the same tracked target as they leave one field of view and enter another.

In effect, the virtual positioning of each device must reflect their real world installation.
Ensure that the Master and Node assignment has been completed before accessing the alignment settings explained here. See the Multi Unit Network Settings for more information.
When devices are correctly aligned and a person walks across multiple devices field of views, there should be one continuous target track.
When devices are NOT correctly aligned a person walking across multiple devices field of views will be shown as multiple target tracks. At the border between neighbouring devices the existing target will drop and a new target will be initialised.
Note looking at the devices pathmap data will also reveal details that indicate a good or bad alignment. See pathmap details here.

 

Device Alignment Basics

Device positioning is done via X & Y coordinates.

The X value corresponds to the X axis, i.e. left-right information, and the Y value corresponds to the Y axis, i.e. forward-back information:

Graph Terminology | Axis, Range & Scale - Video & Lesson Transcript | Study.com

A good way to envisage this is looking down on the area from way above, and through the ceiling, to see the Vectors positioned on a virtual grid.

It is standard practice to leave the master device at position 0, 0 and simply reference the other devices from that one, as this is easiest to administer. Anything to the right of the master will therefore have a positive X value, and anything to the left of the master will have a negative X value.

Alternatively, where a master is in the middle of a group or row of devices, it may be preferable to reference everything from the left most unit, and in that case everything will be positioned with a positive X value.

Similarly all devices in a row will have the same Y value (zero), but adding a positive Y value to a unit means that it is positioned slightly in front of the other devices and a negative Y value means that it is behind.

Essentially how you position devices in the 'virtual world' is up to you and as long as the relative positions of each device is correct then it will work exactly the same.

The X & Y settings are required so that each device knows where it is in relation to all the others. In this way a single person can be correctly tracked across the whole network of units combined field of view. 
It is extremely important that installers, and device configurators, understand that changing the X & Y values will not move devices - only physically relocating devices on the ceiling will do that. X & Y values must reflect how devices have been installed in reality. 

 

Pre-requisites

In all cases devices should be running the very latest firmware version, and importantly master and nodes should be running the same firmware version.

When using the Alignment method detailed below, the firmware of all devices on the wide tracker network should be a minimum of 153.

The very latest version of the RIFT exe or Estate Manager software is also required.

 

How To Configure Device Positions

There are two ways to set the X & Y values of each device. In practice utilizing both methods may be required to position devices in their rough position and then finesse/optimize the positioning afterwards. Listed below.

 

Manually Configuring Device Positions.

To enter the positioning X & Y values manually, first click on the Device icon:

...And a list of all devices connected on the multi-unit network will be displayed:

On this dialog you can change the video view from the master device to one of the nodes, by clicking the yellow 'Play' button.

By clicking the yellow 'Edit' icon you can access the device settings for that device:

Here you can see the settings for the Master device.

It is strongly recommended that the 'Auto Height' button is clicked at this point so that any changes since installation are accounted for. For example, if devices are mounted on ceiling tiles which may have been moved. This will update both the measured height and also the Pitch and Roll angles.

As mentioned at the top of this page, it is recommended that the master is left at X=0 and Y=0 and all other devices are referenced from here

Click the 'Save' button to store any updates to the height and install angles and close this dialog.

 

Now open the settings for the Node:

Note that for any devices that have never been configured with positional information, the X and Y values will be zero by default.
Again, it is recommended that the 'Auto Detect' button is clicked to ensure all height and install angles are correct.

Now enter the X and Y values as required for the installation. 

As you alter the X & Y values and click the Save the virtual setup will change in the main ground plane view:

Remember that you must enter in the actual distances where devices have been installed. The two examples shown above to demonstrate what you'll see as and when you change the values.

In all cases changing the X & Y values will NOT move devices - only physically relocating devices on the ceiling will do that. X & Y values must reflect how devices have been installed in reality. 
To adjust the X & Y values in this way typically requires knowing the values beforehand. I.e. at time of installation the distances are measured as accurately as possible with a tape measure.
It is recommended to move devices and flick between the different video views to see what impact the changes have made. If devices are correctly positioned then the items in the video view that two neighbouring devices can both see will stay in the same relative position as you flick between video views.

Even if the X & Y values have been measured and are known, the second method of setting them is recommended...

 

Configuring Positions Using Auto Alignment Function

As with the manual positing above, you should do the Auto Height function before proceeding. I.e. click the Devices button:

The devices dialog will open:

Now access the settings for each device, one at a time, and click the Auto Height button:

Once all devices are 'Auto Height'-ed, click the 'Alignment' button, shown here:

 

This will change the view slightly, showing the Alignment dialog Window with the live video view now replaced with hires screen grabs of each device:

Note that the screen display might be confusing at first as some screen grabs will be on top of each other hiding parts of others. But don't worry, you can hide each screen grab and/or make them transparent as required, see below.

 

The alignment dialog has the following features:

You are free to adjust the X & Y values as best you see fit, but essentially you must find something in the screen shot of each device and adjust the X & Y values until that thing is in exactly the same place when viewed from either device.

It should be noted that because of the different view points, you need to find something as close to floor level as possible.You cannot use anything too tall, like the top of a table or desk, as each device will see that from a different point of view, and its easy to get confused. But you could use the foot of that table or desk if it is viewable by both devices.
Things like welcome mats, markings on the floor, or in some cases the floor tiles themselves can be used.

In this case we can use the mat on the floor...

By clicking between the master and node views in the Alignment dialog window, we can see that the mats move slightly:

It might be better to adjust the transparency of one of both devices, so as to show both video views at the same time:

The differences can be subtle and the more close to reality the settings are, the less obvious any differences will be. The ultimate aim to to make sure that there are no differences at all, as that means devices are correctly aligned.

Below are the differences between each view with added high lights to illustrate the obvious differences that we can use:

It should be clear there is a slight rotation between devices as well as slight X & Y changes required.

Making changes can be a slow process, so here are the finished results:

Once complete click the Save button to save the X & Y, and yaw angle to the devices.

If you have more than two devices in a configuration, it is recommended to do two neighbouring units at a time. Just click the Show/Hide button as required, to leave the two you want to work on at that time.

Once devices are aligned, be sure to double check that the configured line and/or zone placements are still correct.

 

Things to look for:

In standard view (not in Alignment mode), if a person walks up the overlap between two devices - and splits into multiple targets - then something isn’t right. In these cases, click the ‘Auto Detect’ button to ensure the height, pitch and roll are set correctly first, and then go through this alignment step.

If devices are new, they will start at the default (0,0) position, which means they will be on top of each other on the main view, so start by roughly spreading them out so you can more easily see what you need to do.

Start by working out roughly where devices should be, e.g. does a device belong on the left or right of another device. Do this by flicking between the screen grabs, and looking for common features that indicate where they are in relation to each other.

Note that devices could have been installed at strange angles or the opposite way around from each other, so experiment with the Yaw angle to see if that makes things clearer. Setting one device to 180 degrees will flip its view around completely, for example.

Remember that common features in neighbouring views must be at floor level to be usable for alignment. Look for things like 'Welcome mats', signage/directional arrows on the floor, floor tiles and other patterns, and any other obvious marks.

Note that glare seen on the floor from overhead lighting cannot be used for alignment as each device will see it differently, but shadows common to neighbouring views will be in the same place, and so can be used if required.

Unless you're installing multiple rows of devices covering a large area for dwell applications, the devices will commonly be in a single row, which means the majority of changes required will be to the devices X values. The Y values will be roughly the same.

Once devices are in roughly the right place start fine tuning the values by flicking between neighbouring views and seeing how things change. 

If you have more than two devices next to each other start at one side and work on two at a time. For example if you have three units, hide the right unit so only the left and middle unit are shown and get those aligned. Then hide the left unit and show the middle and right unit, and align those - but remember to only work on the settings of the right unit, or you will mess up the alignment between left and middle unit.