About | Support | Discussion Forum | Index | Service Status | ros @ Robotics Stack Exchange

Only released in EOL distros:   indigo

Overview

anj_featurenav implements a feature-based learning and navigating jockey based on free OpenCV feature descriptor and matcher. The package defines the feature extractor and the descriptor matcher functions required by the featurenav_base package to obtain a working couple learning/navigating jockeys.

Usage

ROS API

The words feature and descriptor are used as synonymous.

Parameters

General

~<name>/feature_detector/type (String, default: "FAST")

• algorithm for feature detection. Can be one of "FAST", "STAR", "ORB", "MSER", "GFTT" (GoodFeaturesToTrack), "Dense", "SimpleBlob". They make use of their corresponding algorithm in the OpenCV library (for example "FAST" implies the use of "cv::FastFeatureDetector")

~<name>/descriptor_extractor/type (String, default: "BRIEF")

• algorithm for feature extraction. Can be one of "ORB", "BRIEF". They make use of their corresponding algorithm in the OpenCV library (for example, "SIFT" implies the use of "cv::BriefDescriptorExtractor")

~<name>/descriptor_matcher/type (String, default: "FlannBased")

• algorithm for feature matching. Can be one of "BruteForce", "FlannBased". They make use of their corresponding algorithm in the OpenCV library (for example, "FlannBased" implies the use of "cv::FlannBasedMatcher")

For the FAST algorithm

~<name>/feature_detector/threshold (Int, default: 1)

• cf. OpenCV documentation http://docs.opencv.org/modules/features2d/doc/feature_detection_and_description.html#fast

~<name>/feature_detector/nonmax_suppression (Bool, default: true)

• cf. OpenCV documentation

For the STAR algorithm

~<name>/feature_detector/max_size (Int, default: 16)

• cf. OpenCV documentation http://docs.opencv.org/modules/features2d/doc/common_interfaces_of_feature_detectors.html?highlight=star#starfeaturedetector

~<name>/feature_detector/response_threshold (Int, default: 30)

• cf. OpenCV documentation

~<name>/feature_detector/line_threshold_projected (Int, default: 10)

• cf. OpenCV documentation

~<name>/feature_detector/line_threshold_binarized (Int, default: 8)

• cf. OpenCV documentation

~<name>/feature_detector/suppress_nonmax_size (Int, default: 5)

• cf. OpenCV documentation

For the ORB algorithm, as feature detector

~<name>/feature_detector/scale_factor (Float, default: 1.2)

• cf. OpenCV documentation http://docs.opencv.org/modules/features2d/doc/feature_detection_and_description.html#orb

~<name>/feature_detector/n_features (Int, default: 500)

• cf. OpenCV documentation

~<name>/feature_detector/n_levels (Int, default: 8)

• cf. OpenCV documentation

~<name>/feature_detector/edge_threshold (Int, default: 31)

• cf. OpenCV documentation

~<name>/feature_detector/first_level (Int, default: 0)

• cf. OpenCV documentation

~<name>/feature_detector/wta_k (Int, default: 2)

• cf. OpenCV documentation

~<name>/feature_detector/score_type (Int, default: 0 (i.e. Harris score))

• 0 for Harris score, 1 for FAST score, cf. OpenCV documentation

~<name>/feature_detector/patch_size (Int, default: 31)

• cf. OpenCV documentation

For the MSER algorithm

~<name>/feature_detector/delta (Int, default: 5)

• cf. OpenCV documentation http://docs.opencv.org/modules/features2d/doc/feature_detection_and_description.html#mser

~<name>/feature_detector/min_area (Int, default: 60)

• cf. OpenCV documentation

~<name>/feature_detector/max_area (Int, default: 14400)

• cf. OpenCV documentation

~<name>/feature_detector/max_variation (Float, default: 0.25)

• cf. OpenCV documentation

~<name>/feature_detector/min_diversity (Float, default: 0.2)

• cf. OpenCV documentation

~<name>/feature_detector/max_evolution (Int, default: 200)

• cf. OpenCV documentation

~<name>/feature_detector/area_threshold (Float, default: 1.01)

• cf. OpenCV documentation

~<name>/feature_detector/min_margin (Float, default: 0.003)

• cf. OpenCV documentation

~<name>/feature_detector/edge_blur_size (Int, default: 5)

• cf. OpenCV documentation

For the GFTT algorithm

~<name>/feature_detector/max_corners (Int, default: 1000)

• cf. OpenCV documentation http://docs.opencv.org/modules/features2d/doc/common_interfaces_of_feature_detectors.html?highlight=gftt#goodfeaturestotrackdetector

~<name>/feature_detector/block_size (Int, default: 3)

• cf. OpenCV documentation

~<name>/feature_detector/quality_level (Float, default: 0.01)

• cf. OpenCV documentation

~<name>/feature_detector/min_distance (Float, default: 1)

• cf. OpenCV documentation

~<name>/feature_detector/k (Float, default: 0.04)

• cf. OpenCV documentation

~<name>/feature_detector/use_harris_detector (Bool, default: false)

• cf. OpenCV documentation

For the Dense algorithm

~<name>/feature_detector/feature_scale_levels (Int, default: 1)

• cf. OpenCV documentation http://docs.opencv.org/modules/features2d/doc/common_interfaces_of_feature_detectors.html?highlight=gftt#densefeaturedetector

~<name>/feature_detector/init_xy_step (Int, default: 6)

• cf. OpenCV documentation

~<name>/feature_detector/init_img_bound (Int, default: 0)

• cf. OpenCV documentation

~<name>/feature_detector/init_feature_scale (Float, default: 1)

• cf. OpenCV documentation

~<name>/feature_detector/feature_scale_mul (Float, default: 0.1)

• cf. OpenCV documentation

~<name>/feature_detector/vary_xy_step_with_scale (Bool, default: true)

• cf. OpenCV documentation

~<name>/feature_detector/vary_img_bound_with_scale (Bool, default: false)

• cf. OpenCV documentation

For the ORB algorithm, as descriptor extractor

~<name>/descriptor_extractor/scale_factor (Float, default: 1.2)

• cf. OpenCV documentation http://docs.opencv.org/modules/features2d/doc/feature_detection_and_description.html#orb

~<name>/descriptor_extractor/n_features (Int, default: 500)

• cf. OpenCV documentation

~<name>/descriptor_extractor/n_levels (Int, default: 8)

• cf. OpenCV documentation

~<name>/descriptor_extractor/edge_threshold (Int, default: 31)

• cf. OpenCV documentation

~<name>/descriptor_extractor/first_level (Int, default: 0)

• cf. OpenCV documentation

~<name>/descriptor_extractor/wta_k (Int, default: 2)

• cf. OpenCV documentation

~<name>/descriptor_extractor/score_type (Int, default: 0 (i.e. Harris score))

• 0 for Harris score, 1 for FAST score, cf. OpenCV documentation

~<name>/descriptor_extractor/patch_size (Int, default: 31)

• cf. OpenCV documentation

For the BRIEF algorithm

~<name>/descriptor_extractor/bytes (Int, default: 32)

• cf. OpenCV documentation http://docs.opencv.org/modules/features2d/doc/common_interfaces_of_descriptor_extractors.html?highlight=brief#briefdescriptorextractor

For the BruteForce algorithm

~<name>/descriptor_matcher/norm (String, default: "L2")

• one of "L1", "L2", "HAMMING", "HAMMING2", cf. OpenCV documentation http://docs.opencv.org/modules/features2d/doc/common_interfaces_of_descriptor_matchers.html#bfmatcher

~<name>/descriptor_matcher/cross_check (Bool, default: false)

• cf. OpenCV documentation

The SimpleBlog algorithm takes no parameter, cf. http://docs.opencv.org/modules/features2d/doc/common_interfaces_of_feature_detectors.html?highlight=simpleblob#simpleblobdetector.

The FlannBased algorithm takes no parameter, cf. http://docs.opencv.org/modules/features2d/doc/common_interfaces_of_descriptor_matchers.html?highlight=flannbasedmatcher#bfmatcher.