robokudo.descriptors.analysis_engines.filereader

Analysis engine for processing stored RGB-D data from files.

This module provides an analysis engine that demonstrates how to read and process RGB-D data stored in the local filesystem. It implements a pipeline for reading Kinect 360 data and performing basic object detection and segmentation.

The pipeline implements the following functionality: - Reading RGB-D data from files - Image preprocessing - Point cloud cropping with restricted field of view - Plane detection - Point cloud cluster extraction

Note

This engine requires the robokudo_test_data repository to be cloned into your workspace. The repository contains example data from a Kinect 360 camera in high-resolution mode.

Warning

Clone https://gitlab.informatik.uni-bremen.de/robokudo/robokudo_test_data before running this pipeline.

Classes

AnalysisEngine

Analysis engine for processing stored RGB-D data.

Module Contents

class robokudo.descriptors.analysis_engines.filereader.AnalysisEngine

Bases: robokudo.analysis_engine.AnalysisEngineInterface

Analysis engine for processing stored RGB-D data.

This class implements a pipeline that reads RGB-D data from files and performs object detection and segmentation. It is designed to work with high-resolution Kinect 360 data stored in the robokudo_test_data repository.

The pipeline includes: - Collection reader for file-based RGB-D data - Image preprocessing - Point cloud cropping (restricted FOV) - Plane detection - Point cloud cluster extraction

Note

The pipeline is configured to loop through the input data continuously and applies specific fixes for Kinect height data.

name()

Get the name of the analysis engine.

Returns:

The name identifier of this analysis engine

Return type:

str

implementation()

Create a pipeline for processing stored RGB-D data.

This method constructs a processing pipeline that reads RGB-D data from files and performs object detection and segmentation. The pipeline is configured to read data from the robokudo_test_data repository.

Configuration details: - Data source: robokudo_test_data/data directory - Kinect height fix mode enabled - Color to depth ratio: 0.5, 0.5 (for high-res mode) - Restricted FOV: -0.3m to 0.3m in X axis

Returns:

The configured pipeline for RGB-D data processing

Return type:

robokudo.pipeline.Pipeline

Warning

Make sure the robokudo_test_data repository is cloned and available in your ROS workspace before running this pipeline.