took place the release of the free library opencv 4.7 (Open Source Computer Vision Library), providing tools for processing and analyzing the contents of the images. OpenCV provides more than 2500 algorithms, both classical and reflecting the latest achievements in the field of computer vision and machine learning systems. The library code is written in C ++ and is distributed under the BSD license. Bindings are prepared for various programming languages, including Python, Matlab and Java.
The library can be used to recognize objects in photographs and videos (for example, recognition of people and figures of people, text, etc.), tracking the movement of objects and cameras and chambers , classification of actions on video, transforming images, extracting 3D models, forming 3D space from the image from stereo chamber, creating high-quality images through a combination of images of lower quality, searching for objects similar to the presented set of elements, the use of machine learning methods, arrangements markers, identifying common elements in different images, automatic elimination of defects, such as the effect of red eyes.
Among the changes in new issue :
- Significant optimization of the productivity of the collapses in the DNN module (Deep Neural Network) with the implementation of machine learning based on neural networks. Implemented an algorithm for fast grape ribbon . Added new layers of onnx (Open Neural Network Exchange): Scatter, Scatternd, Tile, Reducel1 and Reducemin. Added support for the framework Openvino 2022.1 and the backing Cann .
- Improved quality determination and decoding of QR codes.
- Added support for visual markers aruaco and apriltag.
- added the Nanotrack V2 tracker based on neural networks.
- The stackblur blurred algorithm.
- Added support for FFMPEG 5.x and Cuda 12.0.
- A new API for manipulations by multi -page images.
- Added support for the Libspng library for PNG format.
- Libjpeg-Turbo is involved in acceleration using SIMD instructions.
- For the Android platform, support is supported by H264/H265.
- all basic APIs for Python.
- added a new universal backend for vector instructions.