Navigation :

Analytics Settings

Analytivs Overview

Video analytics allow you to select Preset and/or AutoLearn mode for analytics. The parameters defined in the Video Analytics subtab can also determine when video analytics shift from Preset to AutoLearn mode.

For detailed information about analytics, hover your cursor over the round question mark icon in the analytics subtab.

Peset Mode

presets mode requires manual camera calibration, or defining the areas in a camera’s video picture for traffic detection. These detection areas, which are plotted to scale, are depicted by outlines superimposed over the camera video. When presets mode is active, you will see it labeled Active under the center of the video.

AutoLearn Mode

AutoLearn mode is an automatic method of obtaining some traffic incident and congestion data and requires no camera calibration. When AutoLearn mode is activated, you will see it labeled Active at the bottom left of the video.

When AutoLearn and presets modes are both enabled and the camera is running in presets mode, both AutoLearn mode and Presets mode will be shown as Active under the video.

If a camera view is too far off a preset to operate in presets mode, it can also still operate in AutoLearn mode.

View Quality

View quality applies only to cameras operating in AutoLearn mode. View quality refers to how much of the video contains motion compared to areas containing no traffic motion. The more traffic motion consumes the video “screen real estate” in the video, the higher the view quality. If motion portion only takes up a part of the video (i.e., the video contains a high percentage of horizon/sky), the view quality is lower. When view quality is too low, analytics cannot run in AutoLearn mode; however, analytics can still run in presets mode with low view quality.

A camera with view quality below the minimum required for AutoLearn mode can still operate in presets mode.

Drift Tolerance and Drift Correction

Drift tolerance allows you to set a threshold at which the system no longer analyzes the system in presets mode. If the current camera drift is greater than the tolerance level, the system will no longer operate in presets mode. When current camera drift is within the tolerance threshold, you can enable drift correction to enable the system to adjust your preset to match slight drifts of the camera.

Drift Correction allows the software to try and correct the calibrated lane lines to continue getting more robust data. This is useful for cameras that may drift slightly off the calibrated preset view due to wind and other conditions.

A camera view which exceeds specified drift correction limits required for presets mode can still operate in AutoLearn mode unless view quality is too low.

Configuring Analytics

Default Analytics Settings

By default, analytics are enabled when a camera is added to the system. The default settings are as follows:

Enable Analytics is selected
Enable AutoLearn if view quality greater than 15 is selected
Enable Presets is selected
Drift Tolerance is set at 30
Drift correction if below threshold is selected

Changing Analytics Settings

To enable either AutoLearn or presets mode analytics, select the Enable Analytics check box. Deselecting this box disables all analytics for the camera.

To enable AutoLearn mode, select the Enable AutoLearn check box and specify the minimum view quality required for AutoLearn mode.

To enable analytics in presets mode, select the Enable Presets check box.

If you select Enable Presets on the Analytics subtab but you have not calibrated a preset for the camera, analytics will run in AutoLearn mode unless view quality is too low. Analytics may also shift from presets mode to AutoLearn mode if the camera view exceeds the limits set for drift tolerance. For these reasons, enabling presets does not guarantee that presets mode will be active for the camera.

Drift Tolerance

To set the drift tolerance, drag the slider across the Drift Tolerance bar or enter a number in the field to the right of the bar. The current settings for drift tolerance range are indicated by the black brackets on the graph below the Drift Tolerance bar.

The current amount of drift from the active preset camera view is shown in the width of blue field on the graph below the Drift Tolerance bar. It is also displayed under the lower right corner of the video as a numerical value from 1 to 200.

When the current camera view is outside the specified drift tolerance range, the system cannot operate in preset mode but may operate in AutoLearn mode.

Drift Correction

Select Drift correction if below threshold to allow the software to correct the positioning of lane lines when current camera drift is within the specified drift tolerance limits.

Multi-save

You must save your analytics settings by clicking the Save button. To save the same settings to multiple cameras, select the multi-save check box. On the subsequent screen, you will select the cameras to which you want to apply the new settings and click the Save button to save them.

You can use the multi-save function to change the default view quality and drift tolerance settings for the entire system.

Example Analytics Scenarios

Below are images of some possible variations in the Analytics subtab:

Example 1: Using a calibrated preset with drift correction

In this example, the camera is running in presets mode and drift correction is enabled. AutoLearn mode is also enabled and view quality is set at a minimum of 15. The current view quality is 31, and the current degree of drift from the camera preset is 4.

Example 2: Off preset and running in AutoLearn mode

In this example, the camera is running in AutoLearn mode and off preset. Although a preset has been calibrated, the camera view is too far off preset to function in presets mode. Drift tolerance is set at 30, and the current degree of drift is 38. When drift exceeds specified drift tolerance, the analytics revert automatically to AutoLearn mode. View quality is 50 and exceeds the specified minimum of 25, so AutoLearn mode can function.

Example 3: Running in AutoLearn mode, presets not configured

In this example, the camera is running in AutoLearn mode because no preset has been calibrated. View quality is 60 and exceeds the specified minimum of 15. Because drift correction applies only to presets mode, the degree of drift is not calculated nor shown beneath the video.

Example 4: Analytics off, view quality too low and drift too high

In this example, analytics are off because view quality of 18 is below the specified minimum of 25 and the degree of drift is 40, which is higher than the specified threshold. A red notification beside the camera name at the top of the screen indicates that all analytics are off.