VideoCAD

Modeling Multisensor cameras

Hide Navigation Pane

Modeling Multisensor cameras

Previous topic Next topic No directory for this topic No expanding text in this topic  

Modeling Multisensor cameras

Previous topic Next topic Topic directory requires JavaScript JavaScript is required for expanding text JavaScript is required for the print function Mail us feedback on this topic!  

Multisensor cameras allow to cover a large space with high pixel density.

Compared to Fisheye cameras, Multisensor cameras make more efficient use of the available pixels to display the surrounding objects.

In fact, multisensor camera is a bundle of several modules Image sensor + lens. These modules can produce individual images or one common image glued from images from modules.

Many multisensor cameras allow you to direct modules independently, as well as install different lenses on the modules.

 

VideoCAD provides tools for modeling multisensor cameras with independent and linked modules, as well as matrix multisensor cameras with modules located in several lines. Any position of a multisensor camera in space (rotation, tilt, rotation around its axis) can be simulated.

 

See more: Multisensor camera

 

Below we will look at modeling a simple multi-sensor camera with independent modules.

 

 

You can simulate multisensor cameras in VideoCAD in 2 ways:

 

1. Simulation of individual modules of a multisensor camera as separate cameras, combined into the Camera group.

This method is convenient if the modules of the multisensor camera produce individual images and are treated in the project as separate cameras.

Set the Camera group parameter Combined.

You can set the cameras in the group the Multisensor icon or hide icons of all cameras except one.

However, the cameras of the group will remain separate cameras. The combined picture from a group of cameras can be obtained only on the Monitor window or glued manually.

 

2. Using positions for modeling modules of a multisensor camera.

This method is closest to the essence of a multisensor camera.

Positions are copies of camera positions and parameters that are ideal for modeling individual modules.

For even greater convenience, VideoCAD offers a tool for automatic positioning multisensor camera modules.

 

 

Problem

 

There is the AXIS P3717-PLE camera. The camera is installed on the 3m high in the center of a rectangular room 10 * 20 m. It is necessary to control the room in the height range of 0.5-2 meters.

It is required to obtain view area projection, the view area in 3D, distribution of pixel density and image model from this camera.

 

Camera parameters according to the manufacturer's specifications, important for modeling the view area:

Number of modules - 4.
The possibility of free direction of the modules - Yes.
Parameters of each module:

 

Number of pixels - 1920x1080.
Image sensor size - 1 / 2.8 ".
Varifocal lens 3-6mm.

 

Order of work

 

1. Create a camera by clicking the New camera  button. In the camera creation dialog select the Camera icon multisensor.

 

Assigning a special icon is optional. In the future, you can change the icon in the Camera list window.

 

Set camera parameters in the Camera geometry box:

 

Sensor size 1/2,8" 16:9.
Active pixels 1920*1080.
Aspect ratio 16:9.
Lens focal length 3 mm.

 

View area upper bound height - 2 m
View area lower bound height - 0,5 m

 

 

 

2. Model a room 10 * 20 meters with 3 meters of height by the tool Wall. Place 3D models of  windows, doors and people for beauty images.

 

See Example

 

Place the camera in the center of the room.

 

Set the far bound of the view area at the inner edge of the room.

 

Enable modeling shadows by selecting Within projection, in the drop-down menu of the Shadow button.

 

 

 

2. Open the Position of active camera / Multisensor window by the button on the Toolbar.

 

Switch to the Multisensor tab.

 

Mark the Multisensor camera checkbox.

 

Select 4 in the Sensors in line combobox.

 

Select Order of sensor - Tilted. The tilt of the positions is determined by the tilt of the active (parent) camera.

 

Select Angle of sensors 360/4 = 90 degrees to evenly distribute 4 modules in all directions.

 

Select Additional lines (up/down) - 0/0.  No additional lines of sensors.

 

Set the View area upper and lower bounds for the multi-sensor camera.

 

the View area upper and lower bounds for the multi-sensor camera are set independently of the view area bounds of the parent camera, that allows you to construct  projections (sections) of the view area of linked and matrix multisensor cameras in any range of height, regardless of the directions of individual modules.

 

Click the Add positions as sensors button.

 

The active camera will automatically be copied to the created 4 positions.

 

If necessary, you can hide the parent camera by checking the Hide parent camera checkbox.

 

On the projection you can be seen that there are small blind sectors between the positions. They turned out because we did not take into account the distortion of the lenses. View angles of real camera will be slightly larger due to the distortion and the view areas of the modules will intersect without dead sectors.

Camera specification contains values of the real angles so we could model it precisely.

 

Editing position (in space) of a multisensor camera consists of two steps:

1. Editing the position  (in space) of the active camera.

2. Click on the Update sensors button.

 

To edit sensors as positions, switch to the Positions tab.

 

To edit a position:

Copy it to the parent camera by the button POS> CAM.
Edit the camera.
Select the position in the table and copy the camera to the position by the button CAM> POS.

 

When copying a position to the parent (active) camera, the current parameters of the parent camera will be overwritten. If you want to save the parameters, then create an additional hidden position to store the parent camera.

 

 

2. Open the 3D Video window.

 

Right-click on the image in the 3D Video - the Image parameter panel will appear. Switch to the Panoramic tab.

 

Using the switch Positions> Show You can choose to show:

image from the parent camera;
images from positions;
images from the camera and from the positions.

 

In the Images per row combobox , you can choose how many images will be in one row of the multi-image screen. The number of rows is calculated automatically.

 

To see details of the image in real camera resolution:

 

Enable image processing if it is disabled.
Turn on the picture-in-picture (PiP) mode.
Then specify places of interest in the image by clicking by the middle mouse button (wheel).

 

To save the multi-image in real camera resolution of 3840 * 2160:

 

Mark the item in the Main menu (of 3D Video) >Image> Real frame size.
Save the image to a file. Image>Save as.

 

Saving with high resolution may take some time.

 

The model of multisensor cameras allows to simulate high resolution in PiP, images with real camera resolution, motion blur, animated images and camera sensitivity.

You can print multi-images to the PDF report and load to the Monitor window, as well as normal images.

 

 

See also: Fisheye cameras, Positions of active camera