matlab-import-tracking-data

Convert raw CSV, XLSX, and table data into Sensor Fusion and Tracking Toolbox formats.

995|122|Updated Apr 3, 2026
One-click install
npx skills add https://github.com/matlab/matlab-agentic-toolkit --skill matlab-import-tracking-data
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: matlab-import-tracking-data
Source: https://github.com/matlab/matlab-agentic-toolkit/tree/main/skills-catalog/radar/matlab-import-tracking-data
Command: npx skills add https://github.com/matlab/matlab-agentic-toolkit --skill matlab-import-tracking-data

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Recorded trajectory and sensor measurement data rarely matches the input formats expected by MATLAB Sensor Fusion and Tracking Toolbox, forcing users to hand-write conversion code for ground truth timetables, trackingScenarioRecording structs, and objectDetection arrays.

Core Features & Use Cases

  • Truth Data Import: Converts flight logs, GPS logs, ADS-B, and AIS data into trackingScenarioRecording structs, filter tuning timetables, truth logs, or converted tables with coordinate transforms (LLA to ECEF, NED to ECEF).
  • Sensor Data Import: Builds task-oriented dataFormat structs from trackerSensorSpec configurations or legacy objectDetection arrays for radar, IR, lidar, camera, ESM, and sonar measurements.
  • Guided Mapping Workflow: Inspects the user's actual file, infers units and time formats, proposes a column-to-field mapping table, and iterates with the user before generating code.
  • Use Case: A user has a CSV of radar detections with azimuth, range, and Doppler columns and wants to feed them to a tracker; the skill configures a monostatic radar sensor spec, queries its dataFormat, and generates code producing per-timestep structs ready for multiSensorTargetTracker.

Quick Start

Import my flight_log.csv file into a trackingScenarioRecording so I can replay the trajectories in MATLAB.

Frequently Asked Questions about matlab-import-tracking-data

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I import radar detection data into MATLAB for tracking?

Radar detections are imported by configuring a trackerSensorSpec such as trackerSensorSpec('aerospace','radar','monostatic'), querying its dataFormat, and mapping your columns to the required fields. Alternatively, build legacy objectDetection arrays with a standard MeasurementParameters struct when using TOMHT, PHD, or non-EKF filters.

How do I convert GPS or flight log data for Sensor Fusion and Tracking Toolbox?

GPS and flight logs follow the truth pathway: read the file with readtable, convert geodetic coordinates to ECEF using geodetic2ecef, remap platform IDs to sequential integers, and build a trackingScenarioRecording, tuning timetable, or truth log. Time must be zero-initialized and sorted before building the output.

When should I use objectDetection instead of trackerSensorSpec dataFormat?

Use legacy objectDetection when you need TOMHT, PHD, or GridRFS trackers, non-EKF filters like UKF or particle filters, custom measurement models such as TDOA, or when feeding trackingFilterTuner. The task-oriented dataFormat path supports only GNN and JIPDA algorithms with EKF internally.

Can I import raw lidar point clouds or camera images directly?

Raw lidar point clouds need bounding box extraction first, and raw radar point clouds need clustering with dbscan, clusterDBSCAN, or partitionDetections before import. Raw camera images are out of scope because they require an object detector to produce bounding boxes first.

What units and time formats does the import expect?

Target units are degrees for angles, meters for ranges and positions, and m/s for velocities, with conversions inferred from column name suffixes like _deg, _km, or _kts. Time can be POSIX epoch, elapsed seconds, or text datetime, and is always converted to zero-initialized elapsed seconds.