ipa320 / ipa_visual_free_space_detection

free_space_detection

General description

The free_space_detection package

Node: free_space_detection_node

Parameters

m_debug_b (bool, default: false)

Print debug information onto console.

m_extractROI_b (bool, default: true)

Extract lower half of image area as Region of Interest (ROI).

m_printTimings_b (bool, default: false)

Print timing information onto console.

m_writeResults_b (bool, default: false)

Write result images to PNG-files.

m_sigma_gaussian_filter_d (double, default: 0.7)

Sigma of gaussian filter for noise reduction in color image.

m_minH_thresh_d (double, default: 0.3)

Lower boundary for entropy based threshold for hue (H) in H-I-Segmentation.

m_minI_thresh_d (double, default: 0.3)

Lower boundary for entropy based threshold for intensity (I) in H-I-Segmentation.

m_felz_sigma_d (double, default: 0.9)

Sigma parameter for Graph based segmentation of Felzenszwalb and Huttenlocher.

m_felz_k_i (int, default: 200)

K parameter for Graph based segmentation of Felzenszwalb and Huttenlocher.

m_felz_minSize_i (int, default: 250)

MinSize parameter for Graph based segmentation of Felzenszwalb and Huttenlocher, minimum size of segments in pixel.

m_segmentClassification_thresh_d (double, default: 0.6)

Threshold for segment classification: relative amount of free space pixel a region must exceed to be considered as free space.

m_morphSize_i (int, default: 2)

Size of morphological operator used if morphological closing is applied.

m_Vdisp_thresh_d (double, default: 10.0)

Maximum distance (in pixel) for pixel being assigned to line in V disparity image.

m_minValidDisp_thresh_d (double, default: 0.7)

Minimum relative amount of valid disparities in disparity image for disparity information to be considered as reliable.

m_resolution_H_d (double, default: 0.05)

Resolution of hue (H) histogram.

m_resolution_I_d (double, default: 0.03)

Resolution of intensity (I) histogram.

m_Udisp_thresh_d (double, default: 5.0)

Maximum distance (in pixel) for pixel being assigned to line in U disparity image.

m_maxDelay_d (double, default: 450.0)

Maximum time delay in milliseconds allowed between time stamps of color image and disparity image.

m_alpha_d (double, default: 0.7)

Weight for influence of entropy in calculation of entropy based threshold in H-I-Segmentation.

m_maxH_thresh_d (double, default: 0.8)

Upper boundary for entropy based threshold for hue (H) in H-I-Segmentation.

m_maxI_thresh_d (double, default: 0.8)

Upper boundary for entropy based threshold for intensity (I) in H-I-Segmentation.

m_priorityAccuracy_b (bool, default: true)

Perform graph based segmentation. This improves accuracy at object borders but takes much more time.

m_minIntensity_i (int, default: 80)

Minimum mean intensity value for good image quality.

Published Topics

segment_freespace_mask (sensor_msgs::Image)

Result free space mask of the graph based segmentation and classification step.

colorHistogram_freespace_mask (sensor_msgs::Image)

Result free space mask of the H-I segmentation step.

disparity_freespace_mask (sensor_msgs::Image)

Result free space mask of the V disparity step.

final_freespace_mask (sensor_msgs::Image)

Result free space mask.

freeSpacePoints (sensor_msgs::PointCloud2)

Free space 3D points based on final free space mask.

Subscribed Topics

camera_rgb_camera_info (sensor_msgs::CameraInfo)

Camera information of microsoft kinect RGB camera.

camera_rgb_image_color (sensor_msgs::Image)

Color image of microsoft kinect RGB camera.

camera_depth_disparity (stereo_msgs::DisparityImage)

Disparity image of microsoft kinect IR (depth) camera.

camera_depth_points (sensor_msgs::PointCloud2)

3D PointCloud of microsoft kinect IR (depth) camera.

camera_depth_camera_info (sensor_msgs::CameraInfo)

Camera information of microsoft kinect IR (depth) camera.