3.0

Time tracking
Estimated time 67.95 hours
Issues by
android

7/13
build/install

29/37
calibration, 3d

15/18
contrib/text

1/1
contrib/xobjdetect

0/1
core

37/57
documentation

33/37
features2d

16/23
flann

2/2
gpu (cuda)

10/14
highgui-camera

4/7
highgui-gui

6/17
highgui-images

4/6
highgui-video

7/15
imgproc, video

19/27
infrastructure

4/6
ios

5/8
java bindings

8/17
legacy, contrib

4/7
matlab bindings

0/4
ml

7/15
nonfree

0/1
objdetect

7/15
ocl

7/10
photo

1/2
python bindings

11/18
samples

3/7
stitching

1/1
t-api

1/1
viz

2/4

Start date 2014-05-30 Due date 2015-05-15

over 9 years late

63%

262 closed (64%)   145 open (36%)

Related issues
  • Bug #756: MSER detector is missing keypoints on binary images
  • Bug #1244: OpenCV Android native camera doesn't work on some Motorola devices
  • Bug #1334: DynamicAdaptedFeatureDetector
  • Bug #1498: Signficant slow down of SVD in OpenCV 2.3
  • Bug #1603: Build Failure on Mac OS X (GLX support)
  • Bug #1915: avi video without codec leads to crash
  • Bug #1918: Mat.copyTo with mask broken
  • Bug #1929: SURF adjuster crash
  • Bug #1962: CvMLData broken in openCV 2.4
  • Bug #1999: strtod loses all precision after using cv::imshow with OpenCV + QT
  • Bug #2018: DetectionBasedTracker class is not a member of cv namespace
  • Bug #2022: Cannot build snapshot with CUDA on Linux
  • Bug #2067: CascadeClassifier GPU / CPU Detection Mismatches
  • Bug #2138: Memory Leak in imshow (on Windows)?
  • Bug #2144: OpenCV camera crashes during initialization on Android GalaxyS3 (Japanese version SC06D)
  • Bug #2199: Histogram Chi-Square Distance Metric
  • Bug #2235: RTSP over TCP with cvCreateFileCapture not working
  • Bug #2275: Unify interfaces of stereoCalibrate and calibrateCamera
  • Bug #2294: cmake docs target corrupts others (html_docs inclusively)
  • Bug #2359: Native camera does not work on HTC One S with Android ICS 4.0.4
  • Bug #2369: goc2012 video related projects undefined symbols _objc_setProperty_nonatomic
  • Bug #2379: Bug when using convexityDefect
  • Bug #2387: opencv_traincascade "-baseFormatSave" flag for legacy XML file format error
  • Bug #2391: Native camera does not work on HTC One X with Android ICS 4.0.4
  • Bug #2399: Feature Point Region Overlap Computing
  • Bug #2405: Out of bound index in CalcRotation (motion_estimators.cpp)
  • Bug #2419: _InputArray/_OutputArray non-virtual destructors - bad but compatible
  • Bug #2422: <Documentation code sample bug>Histogram Calculation
  • Bug #2427: parallel solvePnPRansac destroys itself
  • Bug #2432: mulSpectrums default function values missing
  • Bug #2471: CommandLineParser crashed
  • Bug #2478: OpenCv2.4.2 fails in cvQueryFrame() for a 1080p video frame
  • Bug #2495: VMWare Fusion Install
  • Bug #2508: solvePnPRansac and solvePnP output consistency
  • Bug #2523: OpenCV 2.4.3 Windows installation error
  • Bug #2530: cvGetWindowHandle with QT windows don't work
  • Bug #2598: cv::dilate doesn't respect image boundaries with shallow copy images.
  • Bug #2599: xcode4.5(4g182) and opencv is the last version,but I can't build
  • Bug #2616: Point_ conversions
  • Bug #2650: SSE resize with AREA interpolation produces incorrect results on 16U input
  • Bug #2676: Failure at Video_calcOpticalFlowDual_TVL1 Regression
  • Bug #2694: window issues on mac os x mountain lion (10.8)
  • Bug #2700: Composite feature extractors and detectors crash on AlgorithmInfo::name()
  • Bug #2717: Wrong index used in Demo Apps for CommandLineParser parser(argc, argv, keys);
  • Bug #2734: solvePnp returns wrong results
  • Bug #2755: Error Handling in Java
  • Bug #2777: OpenCVConfig.cmake should not use include_directories
  • Bug #2817: [OCV 244] - haartraining error ( Assertion failed (elements_read == 1) ... line 1859
  • Bug #2822: cvReleaseCapture do not release the capture device (at least with V4L2 device)
  • Bug #2823: FActory method failure in FeatureDetector
  • Bug #2828: misleading note in doc for VideoCapture::read()
  • Bug #2831: FindChessboardCorners cannot detect chessboard on very large resoluction images
  • Bug #2860: Strange search by docs.opencv.org
  • Bug #2881: error on compiling opencv 2.4.4 .mac lion
  • Bug #2887: inaccurate result in function polyfit
  • Bug #2897: Main module headers are not installed since moved upwards
  • Bug #2899: FileStorage and sprintf
  • Bug #2902: black frame on the depth map with Kineсt
  • Bug #2905: StereoBM.state is not accessible from Java
  • Bug #2924: SparseMat can only support a relatively small number of elements (< 4 194 303)
  • Bug #2927: get NaN-Values from SimpleBlobDetector::findBlobs
  • Bug #2980: opencv-2.4.5 git build broken for grfmt_webp.cpp
  • Bug #2993: Error compiling github version Apr 18, 2013
  • Bug #3007: libv4l backend does not support uncompressed capturing
  • Bug #3009: BAD INSTALL FILES
  • Bug #3013: #include <limits> missing from a few files
  • Bug #3057: findHomography with cv::RANSAC returns empty Mat
  • Bug #3076: matchShapes() always returns 0 , due to HuMoments broken (in 2.4.9)
  • Bug #3077: Hidden attribute on gcc compilation breaks desktop Java bindings
  • Bug #3126: java wrapper can't create cv::Ptr<something> return types
  • Bug #3136: cascade detector returns rectangle outside of image
  • Bug #3184: Missing documentation
  • Bug #3202: Buildbot test results ambiguity
  • Bug #3208: CvBoost train with updating
  • Bug #3215: VideoCapture not working on OSX (relocation of HAVE_QUICKTIME has short-circuited QTKIT)
  • Bug #3225: opencv_ts links in gtest symbols which get pulled in by opencv.pc
  • Bug #3256: .../opencv/modules/core/src/glob.cpp will not compile
  • Bug #3311: cv2.resize causes core dump
  • Bug #3318: Incorrect memory handling in tutorial code (chapter: OpenCV iOS - Image Processing)
  • Bug #3327: Native Camera does not work on Android 4.3 (Nexus 7 2013)
  • Bug #3330: CommandLineParser
  • Bug #3336: ocl::erode, ocl::dilate
  • Bug #3337: ocl::cornerMinEigenVal
  • Bug #3338: ocl::cornerHarris
  • Bug #3340: ocl::adaptiveBilateralFilter
  • Bug #3356: Missing cv::Mat constructor from IplImage* in OpenCV 3.0.0 (git master HEAD)?
  • Bug #3423: decomposeEssentialMat() always return normalized T
  • Bug #3437: Build WITH_OPENGL option not available even if building with QT on Mac OS X
  • Bug #3441: 8-point fundamental matrix estimation seems to be broken
  • Bug #3457: lk_homogrraphy.py sample leaks memory
  • Bug #3460: CV_CAP_PROP'S constants missing from Java bindings
  • Bug #3461: cvTriangulatePoints error
  • Bug #3464: Type errors in Matlab module's bridge.hpp when compiling without some optional modules (calib3d, etc.)
  • Bug #3465: Many type-conversions missing in the Matlab module
  • Bug #3470: memory leak with python
  • Bug #3505: Compilation error due to missing <functional>
  • Bug #3514: Incorrect Chi-Square distance used in LBPH FaceRecognizer
  • Bug #3535: CUDA related compilation issues on MacOSX
  • Bug #3543: Wrong CV_ASSERT when computing descriptors in bagofwords
  • Bug #3565: Portrait orientation on Android
  • Bug #3584: getPolicy method is not threadsafe
  • Bug #3594: Possible cv::divide calculation problem with NaN
  • Bug #3599: cv::FileStorage does not work for std::vector of user-defined struct
  • Bug #3646: Error in "Lucas-Kanade Optical Flow in OpenCV" documentation
  • Bug #3647: "Dense Optical Flow in OpenCV" documentation contains error
  • Bug #3655: CvSVMParams.class_weights not getting set in Python
  • Bug #3672: Releasing VideoCapture / Webcam doesn't work
  • Bug #3681: Android native camera fails on 2.4.9 and 3.0.0
  • Bug #3682: Android native camera fails on 2.4.9 and 3.0.0
  • Bug #3700: Move installed headers to "opencv3/" folder
  • Bug #3703: Downhill Simplex Method
  • Bug #3709: LineSegmentDetector in Python supports only LSD_REFINE_ADV flag
  • Bug #3743: CV_ROW_SAMPLE and CV_COL_SAMPLE not available in Java Machine Learning module
  • Bug #3768: Java binding not working when binding with Python >= 3.3
  • Bug #3786: Don't working haar cascades
  • Bug #3788: Module opencv_cudaoptflow disabled
  • Bug #3790: Documentation: Add documentation for RGBA2* and *2RGBA codes for cvtColor
  • Bug #3791: Build of cudabgsegm fails due to lack of opencv2/legacy.hpp
  • Bug #3803: Errors during building rgbd from opencv_contrib
  • Bug #3809: OpenCV 3.0 Python findContours fails
  • Bug #3825: ocl::HOGDescriptor::detectMultiScale Rect corner points out of range of image
  • Bug #3846: Remove finalize() for all Java classes
  • Bug #3853: Compile error: converting const uchar* to uchar*
  • Bug #3866: ml module is not exported to Python in OpenCV3
  • Bug #3874: Flag CV_PCA_DATA_AS_COL missing in core.hpp
  • Bug #3886: xfeatures2d does not compile without cuda
  • Bug #3891: Mat_ -> Matx conversion sets matrix to all zeros
  • Bug #3895: segmentation fault with Python interface
  • Bug #3897: opencv 3.0 alpha - opencv_adas300(d).lib not generated
  • Bug #3900: BUILD_TBB=ON and BUILD_SHARED_LIBS=ON fail to build
  • Bug #3902: createCLAHE() method is not available in Java binding
  • Bug #3906: Highgui window is not scaled properly when displaying small images
  • Bug #3907: videoio compilation error in cap_msmf.cpp
  • Bug #3928: CvRTrees: setting a random seed
  • Bug #3930: Highgui_Video.prop_fps test fails on Ubuntu 14.04 with FFMPEG
  • Bug #3933: Add README enforcing to all new modules added to contrib repository
  • Bug #3935: cv::pow generates bad results for negative powers of 0
  • Bug #3939: HDR Imaging - assert failure during merge_mertens
  • Bug #3941: opencv_contrib ccv module doesn't build!
  • Bug #3944: CMake: xfeatures2d fails when using world
  • Bug #3946: Color frames in CAP_PVAPI.CPP
  • Bug #3950: using mintheta in houghlines
  • Bug #3963: line_descriptor module fails to build
  • Bug #3970: Fix for CvSVM tutorial documentation
  • Bug #3986: Compiling errors for cap_ffmpeg on Ubuntu 14.10 x64
  • Bug #3988: Three-Dimensional Mats still not accessible in OpenCV Java
  • Bug #3990: Initial conditions ignored when calling DownhillSolver::minimize(x)
  • Bug #3993: CUDA 6.5 NVCUVENC no longer supported
  • Bug #3996: Line Features Tutorial cant be compiled from outside opencv library
  • Bug #3999: Undeclared identifier problem in 3.0.0 alpha
  • Bug #4007: icf train code bug
  • Bug #4011: Different homography results from OpenCV2.x than OpenCV3.0
  • Bug #4017: CvVideoCamera flickering iOS
  • Bug #4019: Bug detected at Mat class constructor #4: Mat::Mat(int rows, int cols, int type, const Scalar& s);
  • Bug #4020: ToneMapDurandImpl::process() fails for images containing intensity zero
  • Bug #4022: No windowedMatchingMask() function in opencv3
  • Bug #4023: OpenCV Vision Challange - Bad link for "EPFL Multi-View Stereo" dataset
  • Bug #4030: AutoTrain fail
  • Bug #4031: Documentation still describes the removed cv::Mat constructor from IplImage*
  • Bug #4033: Activate ml module Python bindings in 3 release
  • Bug #4038: OpenCL kernels for building OpenCV not being generated completely
  • Bug #4042: Download OpenCV
  • Bug #4044: Unable to build with Python 3 support on Ubuntu 14.10
  • Bug #4051: BackgroundSubtractorMOG2 Paramter access denied in opencv 2.4.9
  • Bug #4053: Compile error with FFMPEG
  • Bug #4070: OpenCV 3.0 beta cannot compile with IPP on 32 bit
  • Bug #4077: Backprop implementation is buggy
  • Bug #4081: Erroneous frame seek in VideoCapture::set (CAP_PROP_POS_FRAMES/CAP_PROP_POS_MSEC)
  • Bug #4085: dtree classifier does not work.
  • Bug #4091: OpenCV Video Tutorial bug
  • Bug #4093: Canny can not detect connected contour of black square on a white background.
  • Bug #4101: cvProjectPoints2 - derivative of distortion equation is not correct
  • Bug #4106: Examples runs in debug mode but crash in release mode
  • Bug #4108: Failed linking when ts library enabled on MinGW 4.9.2
  • Bug #4109: python2 binding build error: 'createOptFlow_DeepFlow' : is not a member of 'cv::optflow'
  • Bug #4117: Camera Calibration
  • Bug #4123: cv::ORB works awfully bad on OpenCV 3.0 Beta
  • Bug #4125: Failure to compile with CUDA 7.0 + clang on OSX
  • Bug #4126: typo on “The OpenCV Coding Style Guide”
  • Bug #4129: Algorithm::create doesn't work with [x]features2d modules
  • Bug #4135: XML file can not be loaded.
  • Bug #4136: _OutputArray::assign() not implemented for OutputArray of type MATX
  • Bug #4138: Difference between traincascade software on 2.4 and 3.0 branch
  • Bug #4145: Build fails on Ubuntu 14.04 32-bit (core module)
  • Bug #4151: Mistake in "Contours : Getting Started" of OpenCV-Python Tutorials
  • Bug #4157: CMake inter-target dependency graph contains a cycle
  • Bug #4168: cv::cuda::multiply and cv::cuda::sqr can't handle large vector matrices.
  • Bug #4177: install opencv 3.0 beta with python 3.4 in ubuntu 14.04
  • Bug #4187: Add avfoundation support for normal OSX build, not just iOS
  • Bug #4191: Viz3d crashes when it is closed
  • Bug #4195: ocl_goodFeaturesToTrack: error in maxEigenVal Kernel when using mask
  • Bug #4196: FileNode and FileNodeIterator does not work for std::vector<std::string>
  • Bug #4202: VS 2013 Compilation - meta module commit
  • Bug #4204: what is opencv2/contrib/contrib.hpp
  • Bug #4207: OpenCV 3.0 SVM crashes for large number of iterations
  • Bug #4209: iterating over cv::Mat_ results in infinite loop in OpenCV 3.0.0 (head)
  • Bug #4215: countNonZero() returns wrong value on 32bit machine
  • Bug #4228: Simultaneous Python 2, Python 3 binding
  • Bug #4229: SVM bug in OpenCV 3.0
  • Bug #4239: Missing symbols for cv::rectangle with two points, works fine when using a rect
  • Bug #4243: Change in morphologyEx kernel specifier from opencv-2.4.9
  • Bug #4256: Error in RGB to HLS conversion formula (in cvtColor documentation)
  • Bug #4258: cv::ml::StatModel::calcError not working for responses of type CV_32S
  • Bug #4267: OpenCV 3.0 and Python segfault with detectMultiScale
  • Bug #4274: opencv 2.4.11 imgcodecs not found
  • Bug #4278: size problem in gemm
  • Bug #4279: Algorithm::load() not available for script wrappers
  • Bug #4287: Extra libraries for static linking are not added to pkg-config file
  • Bug #4295: OpenCV 3 RC1 missing files in IOS framework
  • Bug #4296: facedetect.py not working for me .
  • Bug #4297: camera calibration tutorial has wrong radial distortion formula
  • Bug #4298: VideoCapture cannot open video file in Java
  • Bug #4304: Android package does not contain binary libraries for native samples
  • Bug #4305: OpenCV 3.0 RC1 - Python - ShapeDistanceExtractor::computeDistance always returns 0.0
  • Bug #4307: decomposeHomographyMat does not return the expected decomposition
  • Bug #4308: FileStorage cannot write vector<DMatch>
  • Bug #4310: Incomplete documentation for imgcodecs module
  • Bug #4311: Python wrapper warning for win64 platform
  • Bug #4313: Errors in cuda optical flow computation
  • Bug #4314: SVM AutoTrain
  • Bug #4316: Build problem win 64 Visual Studio 2013 (v120) opencv 3.0.0 when BUILD_opencv_world:BOOL=ON
  • Bug #4318: 3.0.0 RC1 build fails on Yosemite, opencl_kernels_core.hpp not found
  • Bug #4319: log function behavior differs from documentation, returns -inf for 0 value elements
  • Bug #4321: SimpleBlobDetector doesn't check when moment m00 is equal to zero
  • Bug #4328: omission of videoio include in documentation
  • Bug #4330: Failure of correctMatches for ideal F matrix
  • Bug #4331: cv::compare(...) : output array type
  • Bug #4333: cv::countNonZero() bug - multi threads access
  • Bug #4334: Poor triangulation for some camera configurations
  • Bug #4335: PCA returns wrong eigenvector
  • Bug #4336: CPU CascadeClassifier with HOG not working anymore with opencv 3.0.0.rc1
  • Bug #4338: Mouse call back always give 0for x and y coordinates in Opengl window
  • Bug #4339: imgcodecs and videoio missing in java docs
  • Bug #4340: push_back() fails to add transposed matrix
  • Bug #4343: cv::cuda::findMinMaxLoc incorrect output for single row matrix
  • Bug #4344: issue #4304 is not fixed
  • Bug #4345: Error in cv::solveLP documentation
  • Bug #4356: Problem in yml parser
  • Bug #4357: OpenCV3 Manager not working on Tegra3 device
  • Bug #4365: 3.0 linux eclipse tutorial points to 2.4 libs
  • Bug #4367: BackgroundSubtractorMOG not working
  • Bug #4369: cv::viz::WCloud can't set PRESENTATION as viz::REPRESENTATION_SURFACE for rendering
  • Bug #4370: Exception in sample stitching_detailedcpp
  • Bug #4372: some sample source codes in documentation need to be revised
  • Bug #4376: Error in Building OpenCV for Microsoft VS 2015
  • Bug #4377: performance issue of UMat in OpenCV 3.0.0-rc1
  • Bug #4379: memory leak in MSER algorithm
  • Bug #4382: cv::Mat memory reallocation using push_back fails
  • Bug #4383: cv::Mat memory reallocation using push_back fails
  • Bug #4385: Segmentation fault with FlannBasedMatcher in python
  • Bug #4388: How to write in OpenCV 3.0
  • Bug #4389: undefined reference to checkIppStatus when compiling with -DWITH_IPP=OFF
  • Bug #4393: Core.multiply crashes JVM
  • Bug #4408: namedWindow(...,cv2.WINDOW_NORMAL) returns object as no attribute error
  • Bug #4416: Missing symbols for cv::rectangle with two points, works fine when using a rect
  • Bug #4417: Bug in conversion from FileNode to Vec2i
  • Bug #4419: DescriptorExtractor.create method fails with 'Specified descriptor extractor type is not supported'
  • Bug #4420: OpenCV Tesseract Wrapper causes memory leak
  • Bug #4422: cv::divide(..) throws "illegal instruction" in OpenCV build in VS2013 made with CMake 3.3
  • Bug #4423: ORB feature detector crashes with AccessViolationException
  • Bug #4424: Cannot find header files when compiling matlab bindings
  • Bug #4431: Open CV copyTo function ( and probably other copying functions) looses precision! (or stops copying after a large number of values)
  • Bug #4432: SURF_CUDA does not match SURF API and is out of date
  • Bug #4438: Errors while building OpenCV 3.0.0 with -D BUILD_TBB flag
  • Bug #4447: Image in sample seamlessclone are missing
  • Bug #4450: VidoeWriter, FeatureDetection like SURF, SIFT etc.. Moments, BOWTrainer, Documentation for Eclipse in Java and so on, are missing in OpenCV 3.0
  • Bug #4451: Unable to link a program in Mac OSX if using cv::CascadeClassifier::detectMultiScale
  • Bug #4452: findContours() doesn't find contours of small areas of binary image
  • Bug #4455: SVM write/load problems with kernels other than linear
  • Bug #4457: Access violation when using sum() on ROI of a Matrix
  • Bug #4459: createTrackbar creates bars out of order and without numbers
  • Bug #4460: RGB2Gray ushort specialization not defined for SSE2 && !SSE4_1
  • Bug #4462: minmaxloc OpenCL kernel uses unaligned memory accesses
  • Bug #4474: findEssentialMat and recoverPose do not work with normalised coordinates as expected
  • Bug #4485: features2d's BRISK, ORB and AKAZE break at detectAndCompute()
  • Bug #4488: Unreasonable code in cv::String::String(const std::string& str, size_t pos, size_t len)
  • Bug #4491: Windows build stiching needs cudawarping
  • Bug #4494: OpenCV Android native camera doesn't work on some Motorola devices
  • Bug #4495: Build Failure on Mac OS X (GLX support)
  • Bug #4496: cameraman.jpg
  • Bugfix #2540: mark multiple return arguments for python bindings
  • Bugfix #2794: AutoBuffer<> ctor compilation error
  • Bugfix #2915: Missing include utility in GPU
  • Bugfix #2953: Fix compiler error due to missing pthread header
  • Bugfix #2954: Fix ocl::Canny accuracy on CPU OpenCL device
  • Bugfix #3530: uninitialized Ptr<CascadeClassifierImpl> ccimpl in bool CascadeClassifier::read(const FileNode &root)
  • Bugfix #3659: videoCapture get documentation
  • Bugfix #3951: [opencv_contrib - opencv_datasets] Compiling errors on Windows
  • Bugfix #3969: Fix for CvSVM tutorial documentation
  • Bugfix #3971: Fix for CvSVM tutorial documentation
  • Bugfix #3972: Fix for CvSVM tutorial documentation
  • Bugfix #3973: Fix for CvSVM tutorial documentation
  • Bugfix #3975: Fix for CvSVM tutorial documentation
  • Bugfix #4000: Undeclared identifier problem in 3.0.0 alpha
  • Bugfix #4001: Declaration file missed in highgui
  • Bugfix #4128: Unicode support in cv::addText via Qt
  • Bugfix #4235: XIMEA Opencv adapter update
  • Bugfix #4269: contrib optflow (sample) motempl - Wrong use of norm
  • Bugfix #4481: Typo in calib3d module doc
  • Bugfix #4482: Typo in calib3d module doc
  • Feature #151: Add analogon to CV_NEXT_GRAPH_EDGE
  • Feature #366: Add SSE optimizations to matrix operations (cvMul, cvDiv)
  • Feature #443: Mat Type Coherence: always returning CV_64FC1 Mat
  • Feature #547: Enhanced matrix Expressions
  • Feature #728: subversion keyword substitution
  • Feature #1093: Add edge aware, weighted debayering algorithms
  • Feature #1221: Add release method to cv::VideoWriter
  • Feature #1268: Unicode Path/Filename for imread and imwrite
  • Feature #1324: always-working video writer
  • Feature #1525: output of GPU detectMultiScale returns multiple of detector size
  • Feature #1683: Create Connected Component routine from generalizing the speckle filter in stereo matching
  • Feature #1754: gpu::compare support for scalars
  • Feature #1849: Add a binary FileStorage format
  • Feature #1857: In user interface, allow user to get the current window positions and sizes
  • Feature #1866: Exposing more Keypoint and Feature2d C++ functions to Java API.
  • Feature #1960: opencv_world TODOs
  • Feature #2007: Make GPU Feature Extractor download/convertKeypoints() static
  • Feature #2013: static (compile-time) assert is needed
  • Feature #2023: No documentation for cv.BoxPoints
  • Feature #2029: Update C++ cheatsheet to the current state
  • Feature #2066: Draw circle grid calibration object
  • Feature #2181: features2d input image conversion
  • Feature #2182: Add DescriptorExtractor::defaultNorm()
  • Feature #2188: Add ASIFT to OpenCV
  • Feature #2206: Reduce SparseMat memory consumption
  • Feature #2269: add new parameter for bilateralFilter (for NaNs in case of 32f)
  • Feature #2299: shift() function
  • Feature #2305: operations on arrays: more cv::Mat return signatures
  • Feature #2306: dense spectral clustering
  • Feature #2308: standardised Distance interface
  • Feature #2309: A big push for sparse support!
  • Feature #2317: implicitly casting away cv::Mat constness
  • Feature #2378: add documentation for command line parser
  • Feature #2402: Remove manual java interface for features2d
  • Feature #2409: CvRTrees member oob_error is not accessible
  • Feature #2424: FileStorage segfaults parsing singleton tags
  • Feature #2442: Finish with migration from http://opencv.willowgarage.com/wiki/
  • Feature #2450: add support convert from RGB to YUV420
  • Feature #2490: cv::imshow() should not be depended on keypress
  • Feature #2517: Video filter slowly on iPhone
  • Feature #2518: Improve readability of InputArray type arg-s in func-s signature
  • Feature #2573: Multiple functions for polygon drawing
  • Feature #2600: cv::CascadeClassifier has no deep copy clone
  • Feature #2625: Const correctnes for cv::VideoCapture::get
  • Feature #2637: Improve Ptr<T>::operator==
  • Feature #2648: Some basic operations on image are slow
  • Feature #2722: Clean up test data
  • Feature #2753: No Tutorial For Visual Studio 2012
  • Feature #2804: Multiple templateMatching
  • Feature #2811: Python 3 support
  • Feature #2862: BoWTrainer interface
  • Feature #2885: add C++ cv::logpolar
  • Feature #2929: Allow histograms (cvCalcHist()) to work with COI and/or iplImageHeader
  • Feature #2998: cv::TriangulatePoints()
  • Feature #3067: Make detectMultiScale return number of neighbors joined into each object
  • Feature #3178: Add EPNP and P3P to cvFindExtrinsicCameraParams2
  • Feature #3218: namedWindow approach in combination with waitKey
  • Feature #3339: ocl::integral with CV_32F, CV_64F
  • Feature #3401: Getting probabilities in CvStatModel
  • Feature #3475: Add ability to serve video stream
  • Feature #3508: Need another calcOpticalFlowPyrLK Java signature taking input List<Mat> pyramid
  • Feature #3548: (Java) Conversion between Mat to IplImage or BufferedImage
  • Feature #3571: Automatically generate C interface to replace legacy API
  • Feature #3576: Request for comments on a render module for OpenCV
  • Feature #3651: proposal for OCL_* performance tests for new T-API functions
  • Feature #3749: CLAHE Lacks Support for Other Image Types
  • Feature #3796: Update Media Foundation back-end for VideoCapture
  • Feature #3848: hogdescriptor::detectmultiscale to output feature vectors for detections
  • Feature #3873: DynamicAdaptedFeatureDetector for image sequences
  • Feature #3890: Subpixel precision by LineSegmentDetector, Image Processing module
  • Feature #3957: Use CommandLineParser in all C++ samples
  • Feature #4021: new type for Mat class.
  • Feature #4024: Niblack's adaptive thresholding
  • Feature #4040: library inclusion
  • Feature #4041: library inclusion
  • Feature #4069: HDR Radiance format .. imread / imwrite
  • Feature #4071: Python fisheye calibration
  • Feature #4100: Adding support for Metal language under iOS platform.
  • Feature #4192: Get accumulator out of hough transform
  • Feature #4222: MSER detector mask ignored (v3.0 beta)
  • Feature #4224: viz::Camera should be usable without vtk
  • Feature #4300: Drawing functions: Support antialiasing for CV_8UC4
  • Feature #4301: signed gradient for HOGDescriptor
  • Feature #4302: opencv 3.0.0 rc1 is support for VS2010?
  • Feature #4346: 2.4 -> 3.0 transition guide suggestion
  • Feature #4362: New ffmpeg build
  • Feature #4366: Support for Android Studio
  • Feature #4434: Update libpng to version 1.6.17
  • Feature #4443: T-Api in Python bindings
  • Feature #4483: Undocumented feature of cv::imshow()
  • Patch #335: Needs workarounds for Sun Studio C++ standard library missing features
  • Patch #1023: Support WebP format
  • Patch #1236: fast connected components implementation (with patch)
  • Patch #1342: Here is code to write matrices to MATLAB
  • Patch #1921: Added gain/gainmode for AVT cameras
  • Patch #1923: RGBA support added to CV_Bayer
  • Patch #2204: Non-maxima suppression
  • Patch #2304: HighGUI module with several controls inside button bar (former Patch #2178)
  • Patch #2421: auto library link for visual studio using #pragma
  • Patch #2465: Support requesting raw file descriptors from V4L2 and DC1394 capture objects
  • Patch #2641: Tesla K20 + printCudaDeviceInfo = -13 cores
  • Patch #2791: alpha channel support for 8-bit tiffs
  • Patch #2809: HighGUI module with several controls inside button bar (former Patch #2178)
  • Patch #3010: A patch to compile OpenCV on Ubuntu 12
  • Patch #3360: svm: return type choice also for the predict on several samples
  • Patch #3394: Compilation Error OpenNI
  • Patch #4498: Documentation patch imread()