Requirements and recommendations for the Abandoned Objects Detection module🔗
Requirements and recommendations for the All objects mode🔗
Image🔗
Static background.
No windows in the frame.
Small quantity of moving objects.
Constant lighting; for example, light fixtures in the premises.
The abandoned objects are not blocked by other objects.
The abandoned object color differs significantly from the color of the background.
Absence or small number of the outstanding small-size objects.
Scene and camera position🔗
Static and reliable camera fixing.
View from the above, or perspective view from the above.
Absence of refocusing and changing image sharpness.
Absence of non-mobile or slow-moving objects in the frame, for example, trees or clerks sitting at their desks. Such objects in the frame may lead to false triggering of the module. At the same time, the presence of a non-mobile object close to the abandoned object may result in the triggering of the module since the abandoned object and the non-mobile object may incorporate into one moving area.
Absence of the light sources slowly changing their position in the frame.
Influence of the module’s settings🔗
Detection zones: if more than a half of the object is in the zone, the object is detected; if not, it is not detected.
Maximum dimensions: if both dimensions (width and height) are smaller than the set maximum dimensions, the object is detected; if at least one of the dimensions exceeds the set limitations, the object is not detected.
Minimum dimensions: if both dimensions (width and height) are larger than the set minimum dimensions, the object is detected; if a t least one of the dimensions is smaller than the set limitations, the object is not detected.
If one of the dimensions of the object is close to the set limitations, the probability of the module’s triggering decreases.
Warning
Regardless of the zone size, the size of the object must be at least 3% of the frame height.
Warning
Since the current version of the module cannot determine the shifts of the objects (insignificant changes of the object’s position), such shifts will result in the triggering of the module.
In order to use the module, it is required to enable and set up the software motion detector, then activate the module itself.
Requirements and recommendations for the Filter by categories mode🔗
Hardware and software🔗
Hardware resources🔗
Tip
To calculate the parameters of the server and client hardware on which you plan to run Eocortex software, use our Calculator.
Using neural networks on the CPU🔗
The module using neural networks that will run on a central processing unit (CPU) requires an Intel CORE I3-10100 or higher processor with AVX support.
Using neural networks on the graphics card🔗
The module using neural networks that will run on a high-performance graphics card (GPU), it is necessary to follow the requirements below:
A processor that supports AVX instructions.
An NVIDIA video card (GPU) with the computation capacity index of at least 6.5 and with at least 4 GB of memory; the parameters and performance of the video card must be similar or better than those of NVIDIA GeForce GTX 1650 Ti model.
Version of the video card driver at least 460.
Swap file at least half of the total RAM size.
Warning
NVIDIA RTX 50 series video cards cannot be used to run Eocortex neural networks.
If the package will be installed on a virtual machine, it may additionally be required to:
Enable support for AVX instructions in the guest machine settings.
Use GRID drivers for GPU virtualization.
Warning
Eocortex must use video cards selected for running neural networks in monopoly mode. It is not allowed to use such card for other applications or tasks that consume GPU resources, including for displaying video. Simultaneous use of a video card for several tasks may lead to incorrect system operation: from analytics performance degradation to server instability.
Warning
When using a video card, the stable operation of the module is guaranteed on operating systems Windows 10, Windows Server 2016, Ubuntu 20.04, as well as on the newer versions of these operating systems.
Warning
On other operating systems (Windows versions 7 and 8, Windows Server versions 2008 and 2012, Debian), issues may arise when trying to use NVIDIA graphics cards. On Windows 8, this is due to the cessation of support for NVIDIA graphics card drivers. On Debian, the problem is due to the high complexity of installing workable versions of video card drivers.
Warning
When upgrading Eocortex to another version, it is necessary to also upgrade the neural networks package to the corresponding version.
Neural networks🔗
Note
If you are having trouble choosing which network package to install, check out this Usage of neural networks article. This article provides an overview and comparison of neural-network packages.
Video stream🔗
The optimal resolution is HD or FullHD.
Frame frequency: no lower than 5 frames per second.
Image🔗
Lighting in the frame should be uniform and constant.
If the camera is installed in front of a bright light source (the sun behind an entrance door, etc.), it is required to adjust the exposure (or brightness) to ensure that the objects in the frame have a natural color (not overexposed or darkened). In this case, it is permissible for the background to be overexposed.
The image must be in color.
The compression quality must not be lower than average. There must be no significant compression artifacts.
The white balance must be properly adjusted.
Scene and camera position🔗
Abandoned objects should be fully visible.
The frame must not contain mirrored surfaces that give reflections: glasses, mirrors, etc.
The camera can be positioned at an angle to the surface. The tilt angle of the camera should not exceed 35°.
Static and secure camera mount.
Warning
It is not recommended to use PTZ cameras.
Object size🔗
For successful object detection, the object's height must occupy at least 8% of the frame height.
Examples🔗
Bag: backpacks, suitcases, bags, etc.
Jerry can: only plastic and metal jerry cans.
Box: only cardboard and paper boxes, e.g. for packing goods.
