Stereo application

This section lists all properties, inputs and outputs of the stereo application.

Properties

distance_enabled

Type:

Bool

Description:

Enable/disable the computation of the distance map.

pointcloud_enabled

Type:

Bool

Description:

Enable/disable the computation of the point cloud.

source_camera_exposure_lock

Type:

Bool

Description:

Lock/unlock the auto exposure control.

source_camera_exposure_min

Type:

UInt64

Description:

The minimum exposure time in microseconds.

source_camera_exposure_max

Type:

UInt64

Description:

The maximum exposure time in microseconds.

source_camera_exposure_compensation

Type:

Float

Description:

The exposure compensation. If < 0, the image becomes darker, if >0 the images gets brighter.

source_camera_gain_min

Type:

Float

Description:

The minimum value of the analog gain.

source_camera_gain_max

Type:

Float

Description:

The maximum value of the analog gain.

source_camera_digital_gain_min

Type:

Float

Description:

The minimum value of the digital gain.

source_camera_digital_gain_max

Type:

Float

Description:

The maximum value of the digital gain.

source_camera_frame_rate

Type:

Float

Description:

The camera’s frame rate.

source_camera_edge_enhance_strength

Type:

Float

Description:

The strength of the edge enhancement.

stereo_matcher_type

Type:

Enum

Values:

Semi-Global Matching, Block Matching, Deep Neural Network

Description:

Defines the stereo matching algorithm.

stereo_matching_image_width

Type:

UInt32

Description:

The width in pixels of the rectified camera. This image of this camera will be used for the stereo matching. Higher values give better results but increase the computation duration.

stereo_matching_image_height

Type:

UInt32

Description:

The height in pixels of the rectified camera. This image of this camera will be used for the stereo matching. Higher values give better results but increase the computation duration.

stereo_matching_image_fov_h

Type:

Float

Description:

The horizontal field of view in degrees of the rectified camera. The range will be mapped to the number of pixels configured by stereo_matching_image_width.

stereo_matching_image_fov_v

Type:

Float

Description:

The vertical field of view in degrees of the rectified camera. The range will be mapped to the number of pixels configured by stereo_matching_image_height.

stereo_target_camera_enabled

Type:

Bool

Description:

Enable/disable the target camera. This reprojects image, distance map and point cloud into a configurable target camera. If disabled, the model of the rectified cameras will be used.

stereo_target_camera_model

Type:

Enum

Values:

EpipolarEquidistant, Equiangular, Equirectangular, Pinhole

Description:

Set the camera model of the target camera.

stereo_target_image_width

Type:

UInt32

Description:

The width in pixels of the target camera.

stereo_target_image_height

Type:

UInt32

Description:

The height in pixels of the target camera.

stereo_target_image_fov_h

Type:

Float

Description:

The horizontal field of view in degrees of the target camera. The range will be mapped to the number of pixels configured by stereo_target_image_width.

stereo_target_image_fov_v

Type:

Float

Description:

The vertical field of view in degrees of the target camera. The range will be mapped to the number of pixels configured by stereo_target_image_height.

stereo_textureness_filter_enabled

Type:

Bool

Description:

Whether or not to apply the textureness filter. Used to “filter out” regions with low optical texture, where the stereo matching algorithm potentially generates wrong results.

stereo_textureness_filter_win_size

Type:

Int32

Description:

The window size of the textureness filter.

stereo_textureness_filter_average_textureness

Type:

Float

Description:

The required average textureness value inside the window.

Inputs

The stereo application has no inputs.

Outputs

image_orig_0

Type:

hs::network::proto::messages::Matrix

Description:

The raw image of camera 0 as UInt8 matrix with three-channels representing the color channels red, green and blue.

image_orig_1

Type:

hs::network::proto::messages::Matrix

Description:

The raw image of camera 1 as UInt8 matrix with three-channels representing the color channels red, green and blue.

image

Type:

hs::network::proto::messages::Matrix

Description:

The image of the target camera as UInt8 matrix with three-channels representing the color channels red, green and blue. If the target camera is disabled, it contains the rectified image of camera 0.

distance

Type:

hs::network::proto::messages::Matrix

Description:

The distance map as single-channel UInt16 matrix representing the distance in millimeters. The matrix is congruent with the image matrix.

pointcloud

Type:

hs::network::proto::messages::Matrix

Description:

The point cloud as three-channel Float matrix representing the 3D point in the camera coordinate system. The matrix is congruent with the image matrix. The channel order is x, y, z and the unit is in millimeters.

image_rect_0

Type:

hs::network::proto::messages::Matrix

Description:

The rectified image of camera 0 as UInt8 matrix with three-channels representing the color channels red, green and blue.

image_rect_1

Type:

hs::network::proto::messages::Matrix

Description:

The rectified image of camera 1 as UInt8 matrix with three-channels representing the color channels red, green and blue.

stereo_frame

Type:

hs::network::proto::messages::MessageMap

Description:

A map containing the image, the distance and the point cloud. The keys are equal to the topic names.

fps

Type:

hs::network::proto::messages::Float

Description:

The internal frame rate.