Versions Compared

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

  1. Detect Pylon

  2. Geolocate Pylon

  3. Send Pylon coordinates to ground station

Task: Module - Task

Blocking Aspects

Remaining subtasks

Target Completion

Command - Send Turn Instruction

Command Module Completion

Author instructions for command module

March 13th

Task: Module - Task

Blocking Aspects

Remaining subtasks

Target Completion

Responsible

Status

Geolocation - Input Conversion

Measurement of camera FOV, Measurement of GPS position

Design approach to problem, measure camera FOV, implement functions

March 20thApril 6th

Andrew and Kevin

Complete

Geolocation - Statistical Analysis → Use weighted mean approach.

None

Get weights (probability of selection) from YOLO model then take weighted average

March 13th

Target Acquisition - Model Training and Saving

Collecting training data, getting decklink setup and working

Get lightbridge running, get GoPro running, setup gopro → lightbridge, lightbridge → decklink and decklink → opencv

March 13th

April 6th

Addesh

Complete

Geolocation - Collinear Points

Review

April 8th

Aryan

In Review

Geolocation - Get Spread logic

Mock function

April 13th

Shrinjay

In Progress

Geolocation - Run Function

Setup run function + tests

April 13th

Shrinjay

In Progress

Command Module - Configure for telemetry use

Review of by Shrinjay and merging command module

Define data source and output for command module, implement

March 13th

Search Program:

  1. Calculate angle to pylon

  2. Steer to pylon

Shrinjay

Complete

Geolocation - Top Box Selection

See line 89 in utils.py, adjust this to only return top box

April 22

Andrew

In Progress

Taxi Program:

  1. Approach Pylon

  2. Identify boxes left to right

  3. Scan QR code, detect correct or incorrect box

  4. Traverse boxes to select correct box

Task: Module - Task

Blocking Aspects

Remaining subtasks

Target Completion

Responsible

Status

Target Acquisition - Box Model verification and deployment

None

Move model to tensorflow

Train model and save as h5 file

March 27thDone.

April 6th

Shrinjay for Review

Awaiting Integration

Distance Measurement

Camera intrinsics data , sync up on others

Pending status update

March 27thApril 13th

Atharva

In Progress

QR Scanning

Camera setup, Correct QR Code

Identify library, identify how we get the correct QR code, implement library with stream from camera

March 27th

Box Traversal

TBD

TBD

TBD - On hold

Deployment and Testing:

Deployment Requirements

...

Vision → GoPro, Lightbridge, DeckLink, GStreamer

...

Communication → PIGO communication through command module

, switch behaviour

April 6th

In Review

Systems:

  • Processing → Connecting models into programs, finishing threadingOperation → Running computer vision program on ground station

Add pipelines between main and geolocation

Requirement

Blocking Aspects

Remaining Subtasks

Completed Subtasks

Vision

GoPro and Lightbridge Posession

Set up GoPro and Lightbridge.

Connect Lightbridge to DeckLink.

Confirm video flow from GoPro → Lightbridge → DeckLink → GStreamer.

Set up computer.

Set up DeckLink.

Communication

Command Module verification and merge

Wrap up command module.

Confirm command module on firmware end is complete.

System test communication.

Command module draft.

Processing

Command Module merge, Geolocation Module completion, Shrinjay’s threading PR (Yes I’m calling myself out)

Date

Responsible

Processing

Distance Module

Add pipeline between main and command module.

Add pipeline between main and command distance module.

Define and implement flight, search and taxi programs.

Module completion.

Threading draft.

Operation

Completion of entire computer vision system

Install and deploy system.

None.

April 13th

Shrinjay

Command - Add error handling to avoid insta-crash

None.

Design error handling system and implement.

April 6th

Andrew

Notes:

Try Catch blocks are very expensive