Permalink
Fetching contributors…
Cannot retrieve contributors at this time
127 lines (113 sloc) 4.71 KB
Changes in 1.3
-------------------------------
* These are by no order of importance:
Webp image format support added
Read image EXIF data
YCrCb color space added
Image Equalize Added
IPython notebook support added (i.e. 'simplecv notebook' from command line)
Resize added
Merge Color Channels Added
Barcode library switched to zbar
drawRectangle added
findCircle added
Whitebalance
Keypoint detection and matching added (SURF, SIFT, and many others, see findKeypoints() docs)
Optical Flow added (see findMotion() docs)
Color Pallette generator added (see getColorPalette() docs)
Skeletonize added
Smart Threshold (a.k.a. Grab Cut) added, see smartThreshold() docs
Smart Find Blobs, findBlobs using Grab Cut, see smartFindBlobs() docs
Rewrote own Threshold function (a.k.a. Binarize), see threshold() docs
FloodFill added
FloodFill to mask added
DFTFilter added
HighPass filter added
LowPass Filter added
BandPass Filter added
Butterworth Filter added
Gaussian Filter added
Unsharpen Filter added
edgeIntersections added
Contour Fitting added
Line Fitting added
Sobel Edge Detection Added
CAMShift Tracking added (see track() docs)
Image area function added
ImageSet's added
Camera.live() function added
External Scanner Support added (see Scanner class docs)
Digital Camera Support added (i.e. point & shoot and SLR cameras)
Stereo Camera Support added
Logging Support added
Feature Overlap (inside, outside, above, below, etc) comparison functions added
Feature.show() added
Feature aspect ratio added
Feature distance functions added
Feature bounding box added
TONS and TONS of documentation added to everything!
A special thanks in this release to anyone externally who has submitted
code. A thanks to all those that made contributions for Google Summer of Code,
or those that help on the support forums. A special thanks to the
Google Summer of Code students of 2012 for all their help in making
SimpleCV more awesome. Sorry this release took so long, as you can see
we have added tons of features, a lot probably not even listed above. (-xamox)
Changes in 1.2
-------------------------------
scale - updated so it can be a value (eg Image.scale(0.5))
binarize - now defaults to Otsu's method rather than binarize 127
blit - overlay images on top of other images with the blit function
start/stop camera - You now have the ability to start and stop a camera object without restarting simplecv
randomized color - set each feature to support auto coloring
stretch filter fixed
foreground/background (codebook) segmentation - Easily compute the difference between foreground and background
use web based flash to acquire images in webpage
bag of features - This is used to detect features and use those features in image classification
shape detectors - Added functions like isCircle(), isSquare(), etc.
better camera support - Now supports HD and fallbacks so if openCV isn't working it tries other methods to work with camera
python imaging library is now required
hue distance - get the distance in hue from current image to another
template matching - This is basic pattern matching
convolution added
integral image - take the derivative of an image
haar wavelet calculations
SVM (binary/multiclass) support added using OrangeSVM
naive bayes (binary/multiclass) support added
decision learning tree (binary/multiclass) support added
random forrest machine learning added
optical character recognition (OCR) support added via Tesseract
ability to load images from URL. i.e. img = Image("http://www.foo.com/bar.png")
pass a feature to crop function
KNN classifier
image segmentation
Changes in 1.1
-------------------------------
All drawing has been moved to pygame-based surfaces in DrawingLayers
text, and more advanced polygon drawing functions
SDL-based display provided by Pygame
python-cvblob has been removed and replaced by BlobMaker
colorspace support for HSV
videostream out for writing AVI files
rotation operations
feature-crop functions
ColorMap for value -> color conversion
ColorModel for training background/foreground segmentation
numerous code adjustments to fit pylint
calibration script for generating calibration
superpacks include freenect and freenect-python drivers
Changes in 1.0
------------------------------
Superpack installers for Mac and Windows
Debian package available for Ubuntu
Jpegstreamer url now accessable via function
Tutorial/Shell fleshed out substantially
Fixed windows-specific bugs in Camera class
Changes in 0.9
-----------------------
Much improved performance in JpegStreamer
Fixed favicon bug
Internal module structure split by class/function
new Image manipulation functions: rotate(), warp(), shear()
added morph operators dialate(), erode()
new shell mode based on iPython invoke with python -m SimpleCV.__init__
fix threads not exiting cleanly bug