Skip to content

Latest commit

 

History

History
243 lines (213 loc) · 8.51 KB

CHANGELOG.md

File metadata and controls

243 lines (213 loc) · 8.51 KB
Raw

Change Log

2.4.1 (2017-06-04)

  • Upstream Kanji fix.
  • Upstream MP SCSI fix.
  • Fix binary ops names.

2.4 (2017-05-30)

  • Implement faster line detection algorithm.
  • Support line segments detection.
  • Support higher FPS on OpenMV 2 and 3.
  • Add data matrix support.
  • Add more small resolutions.
  • Enable UART1 on OpenMV3/M7
  • Enable VSYNC output on IO pin.
  • Fix QR Code bug.
  • Fix UDP recvfrom bug.
  • Minor fixes, typos and docs updates.

2.3 (2017-03-26)

  • Support WiFi Access Point mode.
  • New BAYER/RAW pixel format.
  • Support RGB VGA frames.
  • 1D barcode support using (zbar).

2.2 (2017-02-28)

  • Add Apriltags support
  • Fix OMV3 bootloader LED pins
  • Enable CAN
  • Enable extra MP modules (json, re, zlib, hashlib, binascii, random)
  • Add Pixy emulation.
  • QR Code bug fixes

2.1 (2017-01-21)

  • New keypoints descriptor (ORB).
  • QR decoding via quirc library (https://github.com/dlbeer/quirc)
  • Support load to FB directly in copy_to_fb.
  • Export lens shading function.
  • Add AGAST corner detector.
  • Implement set_gain/exposure/whitebalance functions.
  • Various optimizations and speedups
  • Fix uSD cache issues on M7.
  • Fix broken ADC on M4 and M7.
  • Fix image compress and compressed.
  • Fix ff_wrapper bug.
  • Fix OMV3 LEDs pinout

2.0 (2016-11-04)

Firmware:

  • WiFi driver fixes.

Image processing:

  • Add HoG (not used yet).
  • Add lens correction function.
  • Add clear image for quick testing.
  • Fix template ROI.
  • Switch to FAST-12.
  • Misc fixes to image library.

1.9 (2016-09-20)

Firmware:

  • Initialize RNG when calling randint.

Image processing:

  • Fix and update Kmeans code.
  • Add ellipse masking function.
  • Add face recognition code and example script.
  • Add Hough Transform code and example script.
  • Add Canny edge code and example script.
  • Add Gaussian function for quick testing.

1.8 (2016-08-31)

Firmware:

  • Mainly WiFi driver fixes, more stable streaming, timeouts and better error handling.
  • Fixed FPS slow down in dark images (max FPS reduction is 30FPS)

1.7 (2016-08-25)

Firmware:

  • Update CMSIS, DSP lib and HAL.
  • Adaptive JPEG quality based on JPEG frame size.
  • Improved self-tests on OV7725.
  • New CPU frequency scaling Python module.
  • Allow setting MLX refresh rate and ADC resolution.
  • Use a dedicated JPEG buffer (improves IDE FPS).

1.6 (2016-07-27)

IDE:

  • Add checkbox to disable the framebuffer update

Firmware:

  • Set FB JPEG quality/subsampling based on frame size.

Image processing:

  • Implement windowing.
  • Implement horizontal and vertical binning.
  • Implement optical flow with phase correlation.
  • Implement copy image to framebuffer for testing.
  • Allow ROIs and step in template matching function.
  • Implemented diamond search for fast template matching.
  • Fix bug in integral_image_sq and lookup.
  • Add new smaller resolutions
  • Improved/fixed JPEG code

1.5 (2016-06-01)

IDE:

  • Fix pinout reference image.
  • Fix reset on bootloader (reset cam just before the bootloader runs).
  • Add an option to erase flash filesystem sectors in bootloader dialog.
  • Show color statistics in a message dialog.

Firmware:

  • Update to MicroPython v1.8
  • Change MLX ADC resolution to 18 bits.
  • Fixed GC collect bug (.bss and .data were not scanned, fixed in MP update).
  • Generate a combined (bootloader + app) dfu and binary images
  • Rename firmware images:
  • bootloader.xxx (CDC bootloader images)
  • firmware.xx (main application firmware images)
  • openmv.xx (combined bootloader+firmware images)

Image processing:

  • Allow image line-by-line pre-processing from Python callbacks.

1.4 (2016-05-02)

IDE:

  • Fix text editor undo bug.
  • New bootloader dialog.
  • Fix and update example scripts.
  • Fix preferences dialog.
  • Remove refresh button.

Firmware:

  • Fixed file wrapper initialization bug.
  • New CDC-based bootloader (works on Linux, Windows and OSX)
  • Implement new sensor functions (disable AGC and AEC)
  • Fix WINC bug overriding sent data.

Image processing:

  • Color codes support.
  • New color blob detector.

1.3 (2016-04-07)

IDE:

  • Implement the IDE copy color function.
  • Update examples menu using categories.
  • Fix conflict with PyInstaller scripts.

Firmware:

  • Add initial WiFi (WINC1500) support.
  • Update WINC1500 driver and firmware to 19.4.4
  • Support WINC1500 firmware update from uSD fw image.
  • Improved MLX (FIR) temperature scaling and drawing.
  • Add WiFi examples (mjpeg streamer, NTP, scan, connect and firmware update)

Image processing:

  • Implement AWB/HMirror/VFlip.
  • Implement mean, median and mode filters.

1.2 (2016-03-19)

IDE:

  • About dialog, license and credits.
  • Pin-out image for quick reference.
  • Check for updates on startup.
  • Support older firmware versions.
  • Retry a few times when connecting.
  • Enable/Disable framebuffer JPEG compression.

Firmware:

  • Support the newer OV7725 sensor.
  • Add snapshot timeout to avoid locking the cam.
  • Fix PWM/Servos timer, channels and pin-mappings.
  • Add OpenMV boards configuration files in omv/boards.
  • Support the new MLX90621 sensor and add proper rainbow scaling.
  • Better script handling, and soft reset support.
  • JPEG-compress the framebuffer to lower the bandwidth and fake double-buffering.
  • YUV to Grayscale conversion on the fly.
  • Add sanity checks and more meaningful error messages.
  • Allocate FatFS LFN buffer on stack (frees 255 heap bytes).
  • Move the FatFS and MSC buffers to main RAM (saves heap and allows DMA access).
  • Use DMA for SDIO transfers.
  • Remove framebuffer mutex (IDE reads images before snapshots).
  • Define pin aliases (P0..P8)/
  • Move LCD to built-in module.

Image processing:

  • Improved iris detection.
  • Edge detection, generic convolution, motion detection and GIF support.
  • JPEG compressor optimizations (70ms @QVGA 320x240) faster BinDCT and 2x2 subsampling.
  • Proper JPEG headers for Grayscale images.
  • Bug fixes in old integral image code and a new integral image using a moving window.
  • Set the number of pixels counted in each blob in imlib_count_blobs.
  • Simplify the image descriptor APIs, use a generic image.load/save/match_descriptor functions.
  • Add HQVGA resolution, and special digital effects support.
  • Support higher Grayscale resolutions (up to QVGA) for most algorithms.
  • Image processing functions accept paths to images on uSD.

1.1 (2015-08-15)

  • Rollback to gtksourceview
  • Use MP peripherals
  • Add ABI version and check it in the IDE
  • Add common cascades to the flash
  • Fix changing pixformat bug
  • Fix sensor reset
  • gc/xalloc race bug
  • Fix sensor clock
  • Update to MP 1.4.4
  • Add udev rules help and check for udev file
  • Update USB PID:VID
  • Update inf file
  • Generate Linux/Windows packages
  • Catch and print syntax errors
  • Add colorbar mode function
  • Optimize the IDE (revert to numpy, use timeout_add etc..)
  • Remove obselete #define from mpconfigboard.h
  • Write colorbar test
  • Fix silkscreen
  • Rename Eagle files
  • Move misc image functions to image module
  • Delay sensor init after USB storage to log errors to file
  • Implement get/set pixel
  • Fix push/pop scope (re-init mp before running scripts)
  • Update examples
  • Fix main script FS template in main.c
  • Remove global misc functions
  • Remove lib folder
  • Fix draw_string
  • Disable built-in DFU on Windows

1.0.3-beta (2014-11-15)

  • Binary packaged using py2exe
  • Mixed 32/64 bit Windows installer
  • Fix USB issue on Windows 7 64-bit
  • Enable color-lookup (was disabled in binaries)

1.0.2-beta (2014-11-11)

  • Fixes USB issues on Windows.
  • New MSI package for Windows users
  • Moved all user data are stored in home directory.

1.0.1-beta (2014-11-2)

  • Minor fixes for compatibility with Windows.

1.0.0-beta (2014-10-31)

  • First release.