mirror/opencv
1
0
Fork 0
mirror of https://github.com/opencv/opencv.git synced 2024-11-24 03:00:14 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

Added G-API section for 4.6

Dmtiry Matveev 2022-06-04 21:39:13 +03:00
parent 2a66891099
commit 70fc2b0f1f
1 changed files with 54 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

55
ChangeLog.md

@ -53,7 +53,60 @@ The summer update for OpenCV 4.x has been released.
- G-API module:
- **TBD**
- G-API framework:
- Introduced a Grayscale image format support for `cv::MediaFrame`: [#21511](https://github.com/opencv/opencv/pull/21511);
- Enabeled `.reshape()` support in the CPU backend: [#21669](https://github.com/opencv/opencv/pull/21669);
- Fixed possible hang in streaming execution mode with constant inputs: [#21567](https://github.com/opencv/opencv/pull/21567);
- Introduced proper error/exception propagation in the asynchronous streaming execution mode: [#21660](https://github.com/opencv/opencv/pull/21660);
- Fixed new stream event handling: [#21731](https://github.com/opencv/opencv/pull/21731).
- Fluid backend:
- Fixed horizontal pass in the Resize kernel, fixed Valgrind issues: [#21144](https://github.com/opencv/opencv/pull/21144);
- Extended Resize kernel with F32 version: [#21678](https://github.com/opencv/opencv/pull/21678), added AVX: [#21728](https://github.com/opencv/opencv/pull/21728).
- Enabled dynamic dispatch for Split4 kernel: [#21520](https://github.com/opencv/opencv/pull/21520);
- Enabled dynamic dispatch for Merge3 kernel: [#21529](https://github.com/opencv/opencv/pull/21529);
- Added a SIMD version for DivC kernel: [#21474](https://github.com/opencv/opencv/pull/21474);
- Added a SIMD version for DivRC kernel: [#21530](https://github.com/opencv/opencv/pull/21530);
- Enabled dynamic dispatch for Add kernel: [#21686](https://github.com/opencv/opencv/pull/21686);
- Enabled dynamic dispatch for Sub kernel: [#21746](https://github.com/opencv/opencv/pull/21746);
- Added a SIMD version for ConvertTo kernel: [#21777](https://github.com/opencv/opencv/pull/21777);
- Fixed kernel matrix size for Sobel kernel: [#21613](https://github.com/opencv/opencv/pull/21613).
- OpenCL backend:
- Intel® OpenVINO™ inference backend:
- Fixed NV12 format support for remote memory when OpenVINO™ remote context is used: [#21424](https://github.com/opencv/opencv/pull/21424).
- Implemented correct error handling in the backend: [#21579](https://github.com/opencv/opencv/pull/21579).
- Fixed ngraph warnings [#21362](https://github.com/opencv/opencv/pull/21362).
- OpenCV AI Kit backend:
- Introduced a new backend to program [OpenCV AI Kit](https://store.opencv.ai/collections/usb) boards via G-API. Currently the backend is in experimental state, but allows to build Camera+NN pipeline and supports heterogeneity (mixing with host-side code): [#20785](https://github.com/opencv/opencv/pull/20785), [#21504](https://github.com/opencv/opencv/pull/21504).
- Media integration:
- Enabled GPU inference with oneVPL and DirectX11 on Windows in Intel® OpenVINO™ inference backend: [#21232](https://github.com/opencv/opencv/pull/21232), [#21618](https://github.com/opencv/opencv/pull/21618), [#21658](https://github.com/opencv/opencv/pull/21658), [#21687](21687), [#21688](https://github.com/opencv/opencv/pull/21688). Now GPU textures decoded by oneVPL decoder can be preprocessed and inferred on GPU with no extra host processing.
- Enabled oneVPL support on Linux: [#21883](https://github.com/opencv/opencv/pull/21883).
- Extended GStreamer pipeline source with Grayscale image format support: [#21560](https://github.com/opencv/opencv/pull/21560).
- Python bindings:
- Exposed GStreamer pipeline source in Python bindings: [#20832](https://github.com/opencv/opencv/pull/20832).
- Samples:
- Introduced a pipeline modelling tool for cascaded model benchmarking: [#21477](https://github.com/opencv/opencv/pull/21477), [#21636](https://github.com/opencv/opencv/pull/21636), [#21719](https://github.com/opencv/opencv/pull/21719). The tool supports a declarative YAML-based config to describe pipelines with simulated pre-/post-processing. The tool collects and reports latency and throughput information for the modelled pipeline.
- Other changes and fixes:
- Moved `GKernelPackage` into `cv::` namespace by default, its `cv::gapi::` alias remain for compatibility: [#21318](https://github.com/opencv/opencv/pull/21318);
- Moved Resize kernel from core to imgproc kernel packages for CPU, OpenCL, and Fluid backends: [#21157](https://github.com/opencv/opencv/pull/21157). Also moved tests appropriately: [#21475](https://github.com/opencv/opencv/pull/21475);
- Avoided sporadic test failures in DivC: [#21626](https://github.com/opencv/opencv/pull/21626);
- Fixed 1D Mat handling in the framework: [#21782](https://github.com/opencv/opencv/pull/21782);
- Reduced the number of G-API generated accuracy tests: [#21909](https://github.com/opencv/opencv/pull/21909).
![](images/github2.png)