# opencv 4.14.0 — OpenCV 4.14.0

- Product: opencv (https://whatsnew.fyi/product/opencv)
- Vendor: opencv
- Date: 2026-07-19
- Version: 4.14.0
- Original notes: https://github.com/opencv/opencv/releases/tag/4.14.0
- Permalink: https://whatsnew.fyi/product/opencv/releases/4.14.0

What's New is an index, not a publisher: every entry below links to the vendor's own release notes, which are the authoritative source. Entries are labelled where they are hand-curated sample data, pre-releases, or drawn from a secondary source such as a developer blog.

Reuse: the summaries, labels and curation here are © What's New. Quote freely with attribution and a link back; wholesale republication of the corpus is not permitted — terms: https://whatsnew.fyi/terms. The vendors' own release notes remain their publishers'.

---

- **fixed** — copyTo on empty fixed-type matrices
- **changed** — Improved precision of RotatedRect::points
- **fixed** — UB in cv::error when breakOnError set to true
- **fixed** — cv::mul overflow for U16 type
- **changed** — Revert Disable IPP with AVX512 in cv::compare because of performance regression
- **fixed** — UBSan function pointer type mismatch in countNonZero
- **fixed** — heap-buffer-overflow in YAML parseKey for empty keys
- **changed** — Optimized flip Horizontal
- **added** — AVX VNNI support
- **fixed** — meanStdDev on Windows for ARM
- **added** — NEON implementation for rotate function
- **fixed** — inplace transpose race by enforcing LLSS ordering via local barrier in OpenCL
- **fixed** — incorrect results for in-place flip on strict OpenCL implementations
- **added** — platform-specific SIMD for cv::reduce REDUCE_SUM
- **added** — safe handling of compressed file level extension (.gz[0-9])
- **changed** — Extended AutoBuffer to mimic std::vector
- **fixed** — signed overflow UB in Point/Point3i::dot()
- **added** — SIMD intrinsics support for norm_mask function
- **added** — ARMPL support for DFT Function
- **added** — OpenCL kernel for UMat::diag to avoid aliasing races

*July, 2026*

- Core module:
    - Fixed copyTo on empty fixed-type matrices [#28371](https://github.com/opencv/opencv/issues/28371)
    - Improved precision of RotatedRect::points [#28381](https://github.com/opencv/opencv/issues/28381)
    - Fixed UB in cv::error when breakOnError set to true [#28446](https://github.com/opencv/opencv/issues/28446)
    - Fixed cv::mul overflow for U16 type [#28579](https://github.com/opencv/opencv/issues/28579)
    - Revert "Disable IPP with AVX512 in cv::compare because of performance regression" [#28587](https://github.com/opencv/opencv/issues/28587)
    - Fixed UBSan function pointer type mismatch in countNonZero [#28581](https://github.com/opencv/opencv/issues/28581)
    - Fixed heap-buffer-overflow in YAML parseKey for empty keys [#28620](https://github.com/opencv/opencv/issues/28620)
    - Optimized flip Horizontal [#28614](https://github.com/opencv/opencv/issues/28614)
    - Added AVX VNNI support [#28684](https://github.com/opencv/opencv/issues/28684)
    - Fixed meanStdDev on Windows for ARM [#28739](https://github.com/opencv/opencv/issues/28739)
    - Added NEON implementation for rotate function [#28609](https://github.com/opencv/opencv/issues/28609)
    - Fixed inplace transpose race by enforcing LLSS ordering via local barrier in OpenCL [#28686](https://github.com/opencv/opencv/issues/28686)
    - Fixed incorrect results for in-place flip on strict OpenCL implementations [#28813](https://github.com/opencv/opencv/issues/28813)
    - Added platform-specific SIMD for cv::reduce REDUCE_SUM [#28782](https://github.com/opencv/opencv/issues/28782)
    - Implemented safe handling of compressed file level extension (.gz[0-9]) [#28923](https://github.com/opencv/opencv/issues/28923)
    - Extended AutoBuffer to mimic std::vector [#28909](https://github.com/opencv/opencv/issues/28909) and use it more [#28907](https://github.com/opencv/opencv/issues/28907)
    - Fixed signed overflow UB in Point/Point3i::dot() [#28959](https://github.com/opencv/opencv/issues/28959)
    - Added SIMD intrinsics support for norm_mask function [#28610](https://github.com/opencv/opencv/issues/28610)
    - Add ARMPL support for DFT Function [#28664](https://github.com/opencv/opencv/issues/28664)
    - Implemented OpenCL kernel for UMat::diag to avoid aliasing races [#28897](https://github.com/opencv/opencv/issues/28897)
    - Optimized norm with RVV HAL [#29057](https://github.com/opencv/opencv/issues/29057)
    - Deprecateed MatCommaInitializer_ [#29060](https://github.com/opencv/opencv/issues/29060)
    - Fixed v_matmul/v_matmuladd scalable semantics and expand lane-group test coverage [#29080](https://github.com/opencv/opencv/issues/29080)
    - Siwtched to better Durand-Kerner initialization in solvePoly [#29109](https://github.com/opencv/opencv/issues/29109)
    - Fixed numerical instability and out-of-bounds store in VBLAS SIMD [#29141](https://github.com/opencv/opencv/issues/29141)
    - Added convertScale data type conversions to RVV HAL [#29174](https://github.com/opencv/opencv/issues/29174)
    - Fixed inverted continuity check in cvReshapeMatND() [#29132](https://github.com/opencv/opencv/issues/29132)
    - Fixed memory leak in OpenCL extensions gathering [#28981](https://github.com/opencv/opencv/issues/28981)
    - Added check to reject nd-matrix dim count above CV_MAX_DIM in FileStorage read [#29301](https://github.com/opencv/opencv/issues/29301)
    - Moved IPP math functions from core module to HAL [#29306](https://github.com/opencv/opencv/issues/29306)
    - parallelize sort_ using parallel_for_ [#29192](https://github.com/opencv/opencv/issues/29192)
    - Fixed potential stack overflow in FileStorage [#29348](https://github.com/opencv/opencv/issues/29348)
    - Vectorizeed meanStdDev [#29339](https://github.com/opencv/opencv/issues/29339)
    - FileStorage reads integers above INT_MAX into float/double without truncation [#29368](https://github.com/opencv/opencv/issues/29368)
    - Optimized minMaxI

_[Truncated at 4000 characters — full notes: https://github.com/opencv/opencv/releases/tag/4.14.0]_