Motion detection

Motion detection has revolutionized how we approach surveillance, offering enhanced security and convenience in various settings. Whether you're monitoring your home, office, or public spaces, motion detection technology can provide invaluable insights and alerts, keeping you informed and prepared for any situation.

While professional surveillance cameras with built-in motion detection offer seamless integration and simplified setup, making them ideal for users seeking an all-in-one solution, Banalytics motion detection software analyzes video footage or images to detect motion within a monitored area, making it a versatile option for existing camera setups. With Banalytics, you can empower even the most basic USB or built-in laptop cameras with advanced motion detection capabilities.

Motion detection can be added to any camera with a Capture video task configured. Click the camera title to expand the tasks underneath. Click the Capture video task. Click the Motion detection sub-task in the bottom right corner. After successfully adding the Motion detection task, you can edit its configuration options and run/stop/delete it. 

Description

Restart mode upon catching an error:

• Stop on failure - not restarted until triggered manually.

• Immediately - tries to restart automatically immediately after catching an error.

• 10 sec - tries to restart automatically with a 10-seconds delay after catching an error.

• 30 sec - tries to restart automatically with a 30-seconds delay after catching an error.

• 1 min - tries to restart automatically with a 1-minute delay after catching an error.

The parameter determines the blurring level applied to the image frames or video feed analyzed for motion detection. This setting helps reduce the effects of image noise and other distortions that affect motion detection accuracy. However, setting the blur parameter too high can also reduce the sensitivity to real-motion events.

Dilation is a process that involves expanding or enlarging the detected area of motion to make it more visible or easier to detect.

The dilate size refers to the number of pixels by which the detected area is expanded or dilated. 

Increasing the dilate size can help make the detected motion more visible and easier to distinguish from the background. However, setting the dilate size too large can result in false positives, where the camera detects motion that is not occurring.

The interval at which motion is checked in the image, measured in milliseconds.

The area size, which triggers motion detection. To properly assess the necessary threshold, look at your picture and the object sizes. 

Turn on the flags to Show detections and Show noises. The red squares depict the areas with detected motion, and the yellow areas are areas with noise. Consider the red ones.

Switched on, the setting displays the areas with detected noises—yellow ones. The setting applies only to the Debug mode Off.

Font size of the area labels.

This parameter is explicitly introduced for PTZ cameras that are not supposed to have a static view over time. It turns off the motion detection for the specified period to avoid triggering when a camera moves or the image zooms in or out.

The frequency at which motion detection events are sent. This can be less frequent than detection (Detection timeout), as it's more likely that the same object will continue to move for a second or longer.

The parameter defines how motion detection should be triggered - either upon motion detection or upon detecting a particular object. Without the AI model selected, set the parameter to Motion only.

Here, you can choose an object detection model to trigger motion detection only upon objects of a particular class. Currently, you can use only two available models with limited classes.  Read more here.

Confidence level of the selected object detection model. Below that threshold, motion detection won't be triggered.

NMS stands for Non-Maximum Suppression. It's a technique used in object detection algorithms to filter out redundant or overlapping bounding boxes predicted by the model.

Object classes which should trigger motion detection. All other objects will be skipped.

When switched on, the setting displays the class name above the areas with detected motion—red ones. This setting applies only to the Debug mode Off.

The parameter determines the duration of time during which the camera will ignore further objects for classification after detecting one. Please note that object detection is a resource-consuming process. You can monitor device processing capacity utilization in the System monitor.