Release 3.9.0 - #1941
Merged
Merged
Conversation
Merge main back into develop
* Remove inherited Buffer bindings * Expand new bindings test Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com> * Clean up commented lines --------- Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>
Use variant-based model overloads in Python bindings
Feat/head metadata update
* Add docstrings for SD props not used on RVC2 * Add RVC4 labels * Fix docstring * Fix default value docstring
Fix cam_test.py to support any device, no matter the resolution.
Map depth-to-color interpolation from the fixed minDepth/maxDepth log range instead of each frame's own min/max, so a given depth always maps to the same color instead of flickering as the per-frame range shifts. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
* Remove calibrationHandler from Rectifiacation * Rectification fix reinitialization
Map depth-to-color interpolation from the fixed minDepth/maxDepth log range instead of each frame's own min/max, matching the Python examples, so a given depth always maps to the same color instead of flickering. Also rewrites tof_align.cpp to match tof_align.py: drops the ToFDepthConfidenceFilter block and its extra window, builds the ToF node with CameraBoardSocket::AUTO instead of a hardcoded socket, and adds a --camera left|right flag to pick the align target camera. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
* Rewrite getStereoPairs * Do not consider non-camera features * Sort by baseline length * Remove getCameraZAxisAngle from API * Fix indentations * Require the same sensorname * Require Z axis not being dominant * Improve tests * Check presents of the extrinsics * Fix that we check for A->B and B->A connections --------- Co-authored-by: MaticTonin <matictonin@gmail.com>
Bring all ToF examples (C++ and Python) to a consistent 30 FPS, add RVC2/RVC4 output-queue detection to tof_all_queues.cpp to match the Python version, and remove leftover comments/typos.
Use double-precision seconds converted to nanoseconds instead of truncating to whole milliseconds, matching the exact half-frame-period margin used by the Python tof_align example. Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Improve Depthai Python CI/CD workflow
Remap bboxes when not aligned in SLC
This reverts commit d6d9f3d.
…the functionality
* Add beta namespace and detections filter node * Update ImgDetectionsFilter node to be device runnable, add BETA namespace as a CMake option, default OFF * Update main workflow and python-main workflows to set DDEPTHAI_BUILD_BETA=ON * Add beta ClassificationParser * Add beta ClassificationSequenceParser * Add beta EmbeddingsParser * Add beta KeypointParser * Add beta HRNetParser * Add beta SuperAnimalParser * Add beta LaneDetectionParser * Add beta ImageOutputParser * Add beta MapOutputParser * Add beta MPPalmDetectionParser * Add beta MLSDParser * Add beta RegressionParser * Add beta XFeatMonoParser * Add beta XFeatStereoParser * Add beta SCRFDParser * Add beta YuNetParser * Add beta RFDETRParser * Add beta FastSAMParser * Add beta PPTextDetectionParser * Parsers need to be device nodes and serializable * Add input and output block events * More input output blocks * Bump FW * Clang format * Enable beta building on develop, release and v** branches. Remove requirement for push event * add build_beta as a workflow dispatch * DEPTHAI_BUILD_BETA default is set to ON * Copilot suggestions * Add BetaNode and default to running on host if RVC2 device * Add READMEs * add parser tests * Add parser configs + bindings + runtime config inputs + tests * Remove M_PI constexpr as it was causing windows build errors * Add examples * bump fw * remove regression_parser example as it is not relevant * Bump yunet model size * python * Bump RVC2 and remove roundtrip example * bump fw * Bump FW
Feat/videoencoder improvements
Feature/hil testing overhaul
* Add shared dai::utility colorizeDepthFrame helper and update examples. Introduces dai::utility::colorizeDepthFrame C++ overloads (cv::Mat and ImgFrame) plus Python bindings, supporting min/max depth, colormap and log/linear scaling. Replaces inline cv2.applyColorMap / cv::applyColorMap calls in C++ and Python depth examples with the new helper. * Fix valid pixels collection * Fix neural_assisted_stereo.cpp * Unify logScale * Deal with infinite invalid pixels * Align comments with implementation * Add utility submodule in pybindings and update examples to use it. Examples also dont set default params --------- Co-authored-by: Viacheslav <viacheslav.telelyukhin@luxonis.com> Co-authored-by: aljazkonec1 <konec.aljaz1@gmail.com>
* Normalize dst points based on output size, set image corners in 4pt transform * FW: Update core with 4pt transform fix
* Add maxPoolSize property to ImageManip * RVC4 FW: Resize ImageManip output pool when running out of frames * RVC2 FW: Update core * RVC4 FW: Merge develop * RVC2 FW: Merge develop * FW: Update core * Fix image manip 4pts test
Added LUXONIS_TELEMETRY_CORRELATION_ID env variable support
#1935) * enforced transform validation in stereo_depth_remap.cppand reduced log spam in camera_multiple_outputs.py * updated camera_multiple_outputs.cpp logging to be consistent with python version --------- Co-authored-by: Viacheslav <viacheslav.telelyukhin@luxonis.com>
* asserted that VideoPlayer successfully opens video on init, removed image processing step from DetectionParser segmentation mask test * url correction --------- Co-authored-by: Viacheslav <viacheslav.telelyukhin@luxonis.com>
* Update Bootloader version and add flashing scripts Co-authored-by: AljazD <aljaz.dukic@luxonis.com>
Contributor
|
Important Review skippedToo many files! This PR contains 405 files, which is 305 over the limit of 100. To get a review, reduce the PR to 100 files or fewer by splitting it into smaller PRs or changing its base branch. Upgrade to a paid plan to raise the limit. Usage-priced reviews support at most 300 files. ⚙️ Run configurationConfiguration used: Repository UI Review profile: ASSERTIVE Plan: Pro Plus Run ID: 📒 Files selected for processing (405)
You can disable this status message by setting the Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
9 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Features
depthai_nodesparsersX_LINK_DEVICE_NOT_FOUNDerrors-fflag to enable factory flashingdepthai --flash <device_ip>. By default, the command flashes the user bootloader; use the-fflag to enable factory flashingalphaScalingvalue for controlling the balance between cropping and retaining the full field of viewdai::utility::colorizeDepthFramein C++ anddai.utility.colorizeDepthFramein PythonImgFrame, OpenCV matrices, configurable depth ranges and color maps, and logarithmic or linear scalingsetKeyframeFrequencyhad no effect on RVC4Bug fixes
setMaxPoolSizeto cap the maximum output-pool sizeMisc
getConnectedCameras(CameraSensorType)overload for filtering connected cameras by sensor typecam_test.pynow selects supported resolutions from the connected camera features instead of assuming 1280 x 800background_classandstridesto NN Archive head metadatamodelargumentBuffertimestamp and sequence-number methods instead of declaring duplicate bindings[RVC4] Luxonis OS compatibility
Integration tested with Luxonis OS 1.27.1, 1.30.1, and 1.33.0.