Reverse Move Analysis

  • Changed Hrands video capture code to capture uncompressed video.
  • Ran two test runs with recording uncompressed images: 16_08_03_10_51_39_TargetRun and 16_08_01_15_14_35_TargetRun
  • Analyzed reverse moves with Peter at Caltech 8/8/16
  • Found potential bug in video capture - moves happening between video recording.
  • Changed analysis code "updateMotorMapsFromTargetData" to show correct iteration number.
  • Found issue in MSIM (see email to Hrand) where angles change by large if point lies outside of patrol region.
  • Example is PID 27 Target 121 Iterations 5 and 8 in 16_08_01_15_14_35_TargetRun
  • It turned out that the reason is an old code rule that extends link 2 by a value from the XML file. This value was set to 7 pixels. Hence the strong deviation in angle.
  • Corrective Action: New code first relaxes phi max by 3deg or so and then takes theta angle from centroid beeing radially projected on max radius.

-- JohannesGross - 08 Aug 2016

Edit | Attach | Watch | Print version | History: r3 < r2 < r1 | Backlinks | Raw View | Raw edit | More topic actions
Topic revision: r3 - 2016-08-12 - JohannesGross
This site is powered by the TWiki collaboration platform Powered by PerlCopyright © 2008-2020 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback