Improve documentation and removed obsolete sections. (#844)

Author: jouillet

libraries/robotics-ai-libraries/README.md

@@ -6,7 +6,7 @@ The Robotics AI Libraries provides reusable libraries spanning optimized math ut
 
 | Library | Documentation | Description |
 | ------- | ------------- | ----------- |
-| [FLANN optimized with oneAPI DPC++](flann) | [Link](flann/README.md) | FLANN is a 3-D processing library used to search for nearest neighbors in three-dimensional or even higher dimensional space. |
-| [Motion Control Gateway](motion-control-gateway) | [Link](motion-control-gateway/README.md) | Communication gateway between Real-Time Motion Control system and ROS2(Robot Operating System 2) system. |
-| [ORB Extractor](orb-extractor) | [Link](orb-extractor/README.md) | The ORB feature extractor comprises various computer vision (CV) kernels such as resize, Gaussian, FAST, compute descriptor, and orientation, along with non-CV functions like distribute_keypoints_via_tree. |
-| [PCL optimized with oneAPI DPC++](pcl) | [Link](pcl/README.md) | Point-Cloud Library (PCL) contains common algorithms for 3-D point cloud processing. |
+| [FLANN optimized with oneAPI DPC++](flann) | [Flann readme](flann/README.md) | FLANN is a 3-D processing library used to search for nearest neighbors in three-dimensional or even higher dimensional space. |
+| [Motion Control Gateway](motion-control-gateway) | [Motion control readme](motion-control-gateway/README.md) | Communication gateway between Real-Time Motion Control system and ROS2(Robot Operating System 2) system. |
+| [ORB Extractor](orb-extractor) | [ORB extractor readme](orb-extractor/README.md) | The ORB feature extractor comprises various computer vision (CV) kernels such as resize, Gaussian, FAST, compute descriptor, and orientation, along with non-CV functions like distribute_keypoints_via_tree. |
+| [PCL optimized with oneAPI DPC++](pcl) | [PCL readme](pcl/README.md) | Point-Cloud Library (PCL) contains common algorithms for 3-D point cloud processing. |

libraries/robotics-ai-libraries/flann/README.md

@@ -38,7 +38,7 @@ $ ./flanntest_dp  ../data/data_.txt | tee /tmp/flann_test_cpu.log
 
 FLANN benchmark logs computing latency in milliseconds, for various k and radius values, between Intel® oneAPI DPC++ level0 GPU and plain CPU execution.
 
-```
+```bash
 [----------] 1 test from oneapi_flann_knn_performance_test
 [ RUN      ] oneapi_flann_knn_performance_test.Positive
 data size is 198273

libraries/robotics-ai-libraries/motion-control-gateway/amr/ydlidar_ros2/README.md

@@ -1,13 +1,12 @@
-YDLIDAR ROS2 PACKAGE V1.3.6
-=====================================================================
+# YDLIDAR ROS2 PACKAGE V1.3.6
+
+---
 
 ROS2 node and test application for YDLIDAR
 
 Visit EAI Website for more details about YDLIDAR.
 
-
-Changes to 3rd party source
-=====================================================================
+## Changes to 3rd party source
 
 This work is based off of the open-source
 [YDLIDAR SDK](https://github.com/YDLIDAR/sdk.git) repository.
@@ -18,146 +17,93 @@ The following patches are provided to enhance the YDLIDAR SDK source:
 | -------------------- | ----------------------------- |
 | [patches](patches)   | Wrap angle beyond 180 degrees |
 
-How to build YDLIDAR ros2 package
-=====================================================================
-    1) Clone this project to your ament's workspace src folder
-    2) Running ament to build ydlidar_node and ydlidar_client
-    3) Create the name "/dev/ydlidar" for YDLIDAR
-    --$ cd workspace/ydlidar/startup
-    --$ sudo chmod 777 ./*
-    --$ sudo sh initenv.sh
+### How to build YDLIDAR ros2 package
+
+1) Clone this project to your ament's workspace src folder
+2) Running ament to build ydlidar_node and ydlidar_client
+3) Create the name "/dev/ydlidar" for YDLIDAR
 
-How to run YDLIDAR ros2 package
-=====================================================================
+```bash
+     cd workspace/ydlidar/startup
+     sudo chmod 777 ./*
+     sudo sh initenv.sh
+```
+
+### How to run YDLIDAR ros2 package
 
 1. Run YDLIDAR node and view using test application
-------------------------------------------------------------
-ros2 run ydlidar ydlidar_node
 
-ros2 run ydlidar ydlidar_client
+```bash
+    ros2 run ydlidar ydlidar_node
+    ros2 run ydlidar ydlidar_client
+```
 
 You should see YDLIDAR's scan result in the console
 
-2.Run YDLIDAR node and view using test application by launch
-------------------------------------------------------------
-launch $(ros2 pkg prefix ydlidar)/share/ydlidar/launch/ydlidar.py
+1. Run YDLIDAR node and view using test application by launch
 
-ros2 run ydldiar ydlidar_client or ros2 topic echo /scan
+```bash
+    launch $(ros2 pkg prefix ydlidar)/share/ydlidar/launch/ydlidar.py
+    ros2 run ydldiar ydlidar_client or ros2 topic echo /scan
+```
 
-Configuration
-=====================================================================
+***Configuration***
 path: ~/.ros2/config.ini
 
-Parameters
-=====================================================================
-port (string, default: /dev/ydlidar)
-
-    serial port name used in your system. 
-
-baudrate (int, default: 230400)
-
-    serial port baud rate. 
-
-frame_id (string, default: laser_frame)
-
-    frame ID for the device. 
-
-low_exposure (singleChannel, default: false)
-
-    indicated whether the LIDAR is single communication(S2) lidar. 
-
-resolution_fixed (bool, default: true)
-
-    indicated whether the LIDAR has a fixed angular resolution. 
-
-auto_reconnect (bool, default: true)
-
-    indicated whether the LIDAR auto reconnection. 
-
-reversion (bool, default: false)
-
-    indicated whether the LIDAR data rotation 180°. 
-
-isToFLidar (bool, default: false)
-
-    indicated whether the LIDAR is TOF(TX8) lidar. 
-
-angle_min (double, default: -180)
-
-    Min valid angle (°) for LIDAR data. 
-
-angle_max (double, default: 180)
-
-    Max valid angle (°) for LIDAR data. 
-
-range_min (double, default: 0.08)
-
-    Min valid range (m) for LIDAR data. 
-
-range_max (double, default: 32.0)
-
-    Max valid range (m) for LIDAR data. 
-
-ignore_array (string, default: "")
-
-    Set the current angle range value to zero. 
-
-samp_rate (int, default: 9)
-
-    the LIDAR sampling frequency.
-
-frequency (double, default: 10)
-
-    the LIDAR scanning frequency.
-
-
-
-
-
-
-
-Upgrade Log
-=====================================================================
-
-2019-12-23 version:1.4.4
-
-   1.support all standards lidar
-
+***Parameters***
 
-2018-07-16 version:1.3.6
+| Parameter           | Description                                                                 |
+|---------------------|-----------------------------------------------------------------------------|
+| `port` (string, default: `/dev/ydlidar`)           | Serial port name used in your system.                              |
+| `baudrate` (int, default: `230400`)                | Serial port baud rate.                                              |
+| `frame_id` (string, default: `laser_frame`)        | Frame ID for the device.                                            |
+| `low_exposure` (singleChannel, default: `false`)   | Indicates whether the LIDAR is single communication (S2) LIDAR.     |
+| `resolution_fixed` (bool, default: `true`)         | Indicates whether the LIDAR has a fixed angular resolution.         |
+| `auto_reconnect` (bool, default: `true`)           | Indicates whether the LIDAR auto reconnects.                        |
+| `reversion` (bool, default: `false`)               | Indicates whether the LIDAR data is rotated 180°.                   |
+| `isToFLidar` (bool, default: `false`)              | Indicates whether the LIDAR is a TOF (TX8) LIDAR.                   |
+| `angle_min` (double, default: `-180`)              | Minimum valid angle (°) for LIDAR data.                             |
+| `angle_max` (double, default: `180`)               | Maximum valid angle (°) for LIDAR data.                             |
+| `range_min` (double, default: `0.08`)              | Minimum valid range (m) for LIDAR data.                             |
+| `range_max` (double, default: `32.0`)              | Maximum valid range (m) for LIDAR data.                             |
+| `ignore_array` (string, default: `""`)             | Set the current angle range value to zero.                          |
+| `samp_rate` (int, default: `9`)                    | The LIDAR sampling frequency.                                       |
+| `frequency` (double, default: `10`)                | The LIDAR scanning frequency.                                       |
 
-   1.Update SDK verison to 1.3.9
- 
-   2.remove imu sync.
+#### Upgrade Log
 
-2018-07-16 version:1.3.5
+#### 2019-12-23 version:1.4.4
 
-   1.Update SDK verison to 1.3.6
+- support all standards lidar
 
-   2.add imu sync.
+#### 2018-07-16 version:1.3.6
 
-2018-04-16 version:1.3.1
+- Update SDK version to 1.3.9
 
-   1.Update SDK verison to 1.3.1
+- remove imu sync.
 
-   2.Increase sampling frequency,scan frequency setting.
+#### 2018-07-16 version:1.3.5
 
-   3.Unified coordinate system.
+- Update SDK version to 1.3.6
 
-   4.Repair X4,S4 LIDAR cannot be opened.
+- add imu sync.
 
-   5.Increased G4 G4C F4Pro LIDAR power-off protection.
+#### 2018-04-16 version:1.3.1
 
-   6.Increased S4B LIDAR low optical power setting.
+- Update SDK version to 1.3.1
 
-   7.Fix the wait time for closing ros node.
-   
-   8.Compensate for each laser point timestamp.
+- Increase sampling frequency,scan frequency setting.
 
-   9.Unified profile, automatic correction lidar model.
+- Unified coordinate system.
 
+- Repair X4,S4 LIDAR cannot be opened.
 
+- Increased G4 G4C F4Pro LIDAR power-off protection.
 
+- Increased S4B LIDAR low optical power setting.
 
+- Fix the wait time for closing ros node.
 
+- Compensate for each laser point timestamp.
 
+- Unified profile, automatic correction lidar model.

libraries/robotics-ai-libraries/orb-extractor/README.md

@@ -1,123 +1,22 @@
-# ORB Feature Extractor - applications.robotics.mobile.orb-extractor
+# GPU ORB Extractor Feature Overview
 
-Description
------------------------------------------------------------------------------
-The ORB feature extractor sample application demonstrates its functionality on both CPU and GPU. The ORB feature extractor is based on the OpenVSLAM version.
+This repository provides tutorials and usage guides for the GPU ORB Extractor Feature, supporting both OpenCV-based and OpenCV-free workflows. The GPU ORB Extractor library enables fast, thread-safe keypoint extraction for visual SLAM and computer vision applications, with support for multiple cameras and multi-threaded processing.
 
-https://github.com/OpenVSLAM-Community/openvslam
+## Features
 
-The ORB feature extractor comprises various computer vision (CV) kernels such as resize, Gaussian, FAST, compute descriptor, and orientation, along with non-CV functions like distribute_keypoints_via_tree.
+- High-performance GPU-accelerated ORB keypoint extraction
+- Supports both OpenCV (`cv::Mat`, `cv::KeyPoint`) and OpenCV-free data structures
+- Thread-safe API for multi-camera and multi-threaded use cases
+- Flexible configuration for single or multiple ORB extractor objects
 
-All the CV-related kernels are offloaded to the Intel GPU using the oneAPI Level Zero interface, and the GPU kernels are written using C-for-Metal. Non-CV functions run on the CPU.
+## Quick Start
 
-For more information about the oneAPI Level Zero interface, refer to the link below.
+Refer to the GPU ORB Extractor Feature Overview page for more details on the pre-requisites.
+[ORB Extractor Overview](https://docs.openedgeplatform.intel.com/edge-ai-suites/robotics-ai-suite/main/robotics/dev_guide/tutorials_amr/perception/orb-extractor/package-use.html)
 
-https://spec.oneapi.io/level-zero/latest/index.html
+## Documentation
 
-For more informaton of C-for-Metal, refer to link below.
-
-https://01.org/c-for-metal-development-package
-
-
-System Requirements for building and running the sample
------------------------------------------------------------------------------
-
-### Require Ubuntu 20.04
-Follow the link below.
-https://ubuntu.com/tutorials/install-ubuntu-desktop#1-overview
-
-### Kernel version 5.8.0 or greater
-
-#### To install latest Ubuntu linux kernel
-
-```bash  
-sudo apt-get upgrade linux-image-generic 
-```
-
-#### To install a specific kernel version (alternatively)
- ```bash
-wget https://raw.githubusercontent.com/pimlie/ubuntu-mainline-kernel.sh/master/ubuntu-mainline-kernel.sh
-```
-```bash
-chmod +x ubuntu-mainline-kernel.sh
-```
-```bash
-./ubuntu-mainline-kernel.sh -i <kernel version to install>
-```
-
-### Install build tools
-```bash
-sudo apt install build-essential cmake
-```
-
-### Install Level Zero runtime
-Follow the link below.
-https://dgpu-docs.intel.com/installation-guides/ubuntu/ubuntu-focal-legacy.html
-
-### Install OpenCV 4.2
-The library only depends on OpenCV for some of cv types like cv::KeyPoint, cv::Mat,
-and function like cv::getGaussianKernel.  
-
-Follow the link below.
-https://linuxize.com/post/how-to-install-opencv-on-ubuntu-20-04/
-
-### Install C for Metal Development Package
-```bash
-wget https://01.org/sites/default/files/downloads/cmsdk20211028.zip
-unzip cmsdk20211028.zip
-cd cm_sdk_20211028
-source setupenv.sh
-```
-
-### Build the library
-```bash
-git clone https://github.com/open-edge-platform/edge-ai-libraries -b release-1.2.0
-cd edge-ai-libraries/libraries/orb-extractor
-mkdir build
-cd build
-cmake -DBUILD_TARGET_PLATFORM=<platform name> -DCMAKE_BUILD_TYPE=<Release/Debug> -DCMAKE_INSTALL_PATH=<path to installation folder> ../ 
-# For ex: use cmake -DBUILD_TARGET_PLATFORM=tgllp -DCMAKE_BUILD_TYPE=Release ../ to build for tigerlake
-# or cmake ../ to build for current platform
-make -j8
-```
-### Install library
-
-***Note*** use sudo if installing into default location of /usr. Otherwise, not recommended.
-```bash
-sudo make install
-```
-### Build the sample
-```bash
-cd ../samples
-mkdir build
-cd build
-cmake ../
-make -j8
-```
-### Run the sample
-```bash
-./feature_extract
-```
-### Build unit tests
-```bash
-# Build the library
-cd ../tests
-mkdir build
-cd build
-cmake ../
-make -j8
-```
-### Run the test
-```bash
-./gaussTest
-./resizeTest
-./fastTest
-./orbdescTest
-./stereoTest
-```
-
-### Adjusting for latency and CPU util
-```bash
-export CPU_SLEEP_TIME=<sleep duration in microseconds>
-```
-***Note*** Higher sleep value leads to higher latency & reduced CPU util and vice versa
+- [API Usage](https://docs.openedgeplatform.intel.com/edge-ai-suites/robotics-ai-suite/main/robotics/dev_guide/tutorials_amr/perception/orb-extractor/api-use.html)
+- [OpenCV-free Usage](https://docs.openedgeplatform.intel.com/edge-ai-suites/robotics-ai-suite/main/robotics/dev_guide/tutorials_amr/perception/orb-extractor/orbocvfree-use.html)
+- [Limitations](https://docs.openedgeplatform.intel.com/edge-ai-suites/robotics-ai-suite/main/robotics/dev_guide/tutorials_amr/perception/orb-extractor/limitation.html)  
+For more details, see the GPU ORB Extractor [index page](https://docs.openedgeplatform.intel.com/edge-ai-suites/robotics-ai-suite/main/robotics/dev_guide/tutorials_amr/perception/orb-extractor/index.html).

libraries/robotics-ai-libraries/orb-extractor/README.md.external

@@ -86,4 +86,3 @@ Additional option
 ### Adjusting for latency and CPU util
 export CPU_SLEEP_TIME=<sleep duration in microseconds>
 # Higher sleep value leads to higher latency & reduced CPU util and vice versa
-