Release highlights:
- Enabled runtime dispatched optimizations for larger set of functions in core and imgproc modules.
- dnn module has got several improvements:
- Inference Engine backend has been switched to NN Builder API, supported Intel® Neural Compute Stick 2
- Reduced peak memory consumption and supported multiple new networks from TensorFlow
- Android Media NDK API support has been added to videoio module
- Added several Perceptually Uniform Sequential colormaps
- New Image Quality Analysis module (quality) has been added to the opencv_contrib, referenceless BRISQUE algorithm has been implemented as well as PSNR, SSIM and others
- Robust Local Optical Flow algorithm has been added to opencv_contrib/optflow module
- Quasi Dense Stereo matching algorithm has been implemented in opencv_contrib/stereo module
- Added Hand-Eye Calibration methods
More details can be found in the Changelog
Most of bugfixes and improvements have made their way to both 3.4 and master branches.
Contributors
Big thanks to everybody who contributed (here is the incomplete list of patch authors; please, report if you contributed but do not see your name here):
opencv
Alexander Alekhin, Dmitry Kurtaev, Vitaly Tuzov, Maksim Shabunin, Namgoo Lee, Rostislav Vasilikhin, Quentin Chateau, Sayed Adel, smirnov-alexey, Alexander Nesterov, Alexey Smirnov, LaurentBerger, Giles Payne, Pavel Rojtberg, Andrey Golubev, Brad Kelly, Liubov Batanina, Ruslan Garnov, berak, catree, Christoph Rackwitz, Kohei Yoshida, Lee Jaehwan, Pedro Ferreira da Costa, Stefan Dragnev, Suleyman TURKMEN, Wu Zhiwen, klemens, npochhi, zuoshaobo, 103yiran, AKAMath, Adrian Renner, AnastasiaaSenina, Arnaud Brejeon, Augusto Fraga Giachero, Ayush Pandey, BALACHANDAR S, Bharat123Rox, CUI Hao (Lab), Christopher N. Hesse, David Carlier, Easton Liu, ExpandingDev, George Wilson, Gregor Mittag, Hannah McLaughlin, Jaewoo Song, Jim Zhou, JoeHowse, Josh Veitch-Michaelis, Juha Reunanen, Kangning Song, Lindsay Roberts, Lubov Batanina, Lucas Towers, Matthew K. Gumbel, Nicolae Rosia, Oleh, Otto Seiskari, Pierre Chatelier, RAJKIRAN NATARAJAN, Rajkiran Natarajan, Ramez Rafla, Raphael Graf, Richard Veale, Scott Ramsby, Thang Tran, Thomas Peters, WuZhiwen, atinfinity, cftang, iPanda, mehlukas, rpici, winice, x3lif
opencv_contrib
Namgoo Lee, Pavel Rojtberg, Alexander Alekhin, clunietp, Hamdi Sahloul, Maksim Shabunin, Raphael Graf, Charles Taylor, Dimitrios Psychogyios, Dmitry Kurtaev, JoeHowse, Julian Seuffert, Richard Veale, Sreejith S, Ted Steiner, Tomoaki Teshima, berak, iindovina, kushashwa ravi shrimali, shimat, tsenst, utkarshg6