Skip to content
This repository has been archived by the owner on May 20, 2024. It is now read-only.

Latest commit

 

History

History
43 lines (37 loc) · 2.46 KB

README.md

File metadata and controls

43 lines (37 loc) · 2.46 KB
Raw

2018-Vision

To play around with python code:

Copy the contents of ArgosTargetDetection.py to Jevois Sandbox module (pythonSandbox.py) until a separate module is built -parameters to tune -HUE_MIN, HUE_MAX (ideal 30 to 80) -SAT_MIN, SAT_MAX -VAL_MIN, VAL_MAX -AREA_THRESHOLD -MIN_ASPECT_RATIO_THRESHOLD

To run the ArgosVsion module in C++:

  • copy src folder contents into src folder of the newly created module.

  • To create a module get the jevois base repos from: https://github.com/jevois/jevois.git and https://github.com/jevois/jevoisbase.git and https://github.com/jevois/jevois_sdk_dev.git

  • Then open a terminal and do: ~/path_to_jevois_repos/jevois/scripts/create_new_module.sh Argos ArgosVision

  • Now copy the src contents into newly created modules src folder (overwrite)

  • In order to compile one MUST have 16.04 or 17.10 (17.04 is no longer supported from February 2018).

  • IMPORTANT: before compiling follow the instruction from https://jevois.usc.edu/ and get the appropriate dependencies

  • now navigate to base folder of the module and run: ./rebuild-platform.sh to compile for ARM(jevois).

  • After compiling copy the jvpkg folder to packages folder of Jevois camera(microSD) and restart.

  • parameters can be found in script.cfg file of the module.

  • As of 02/10/2018 all parametrs in the code are made configurable and can be modified via serial commands on the fly.

Running the module

  • video mapping - 320x254
  • algorithm modes can be changed by sending the following serial command: setpar detect_mode 1
  • detect_modes: 1 - visionTargets, 2 - platforms, 3 - powerCubes
  • following is the structure of the serial message sent out by the module:

| || || || || || || || || || || || | | x || y || d || x || w || h || 0 || 0 || 0 || 0 || 0 || LABEL | | || || || || || || || || || || || |

x - horizontal offset from the center of the frame to center of the target y - vertical offset from the center of the frame to center of the target d - approximate distance in inches from camera to the target w - width of the detected target in pixels h - height of the detected target in pixels LABEL - Target description