QUARC 2022
These release notes describe the new features and changes introduced in QUARC 2022. They are divided into the sections enumerated below.
New Software Compatibility
QUARC 2022 has introduced compatibility with the following third-party software.
Compatibility with MATLAB, Simulink, and Simulink Coder R2022a has been added.
Compatibility with Microsoft Visual Studio 2022 Professional and Community Editions has been added.
New Hosts
QUARC 2022 now supports the following new hosts.
Compatibility with the Microsoft 64-bit Windows 11 operating system has been added. Note that MATLAB R2021b Update 1 and above is needed to support Windows 11.
New Features
The new features introduced in QUARC 2022 are mentioned below.
A QUARC tab is added to the Simulink toolstrip. The tab includes all the common actions that are usually performed in Simulink models. This also replaces the QUARC menu item.
The QUARC tab is only supported in MATLAB R2021b and newer versions.
New Cards Supported
The full list of all the data acquisition cards supported by QUARC 2022 is given by the following link. QUARC 2022 may now also interface to the cards listed below.
Added support for the Quanser Aero 2 device.
Added support for the Quanser QBot 3 device.
The National Instruments PCIe-6343 is an X-Series card for PCIe slot. Please see National Instruments' website for the (PCIe-6343) for the card data sheet.
New Blocks
New blocks have been added to the QUARC Targets library in QUARC 2022. The new blocks are listed below.
Devices Blocks
The Ranging Sensor Simulation block allows a ranging sensor to be simulated using a QUARC model.
Image Processing Blocks
Added a LIDAR Scan Match block which finds the pose of one 2D LIDAR scan relative to another (translation and rotation). It is typically used for localization when LIDAR data is available. The Ranging Sensor block may be used to access LIDAR devices such as the RPLIDAR A2.
The Image Align block aligns a depth image to another image. It is typically used to align a depth image to an RGB image from the same RGBD camera so that depth information can be determined for each pixel of the RGB image.
Multimedia Blocks
The Video3D Extrinsics block gets the extrinsic transformation from one Video3D image sensor to another.
Improved Features
Some of the features improved in QUARC 2022 are enumerated below. There are a great many miscellaneous improvements that have not been listed. Only some of the highlights are listed below.
The LCD Display block has been enhanced to support ST7066U, which is used on Quanser QBot 3.
The Colorization algorithm quantizes the input image (typically a depth image) and maps the quantized depth to colors. The algorithm was modified slightly so that the mapping from quantized depth to color is invertible. Hence, it is now possible to invert the colorized image to get back the quantized depth image using the new Decolorization algorithm. Not all color themes may be decolorized.
The Colorization algorithm for the double data type was optimized to yield a small performance improvement.
A new color theme called Hue was added to the list of color themes. It is intended to match the algorithm used by the following article: Depth Image Compression by Colorization, although the default color theme, KRYGCBMW, and its permutations are better because they provide more effective bits when quantizing the depth image.
A new Decolorization algorithm was added to the Image Transform block that takes a depth image colorized by the Colorization algorithm and returns the original quantized depth image. Note that, in general, only a quantized version of the depth image is recoverable because the effective number of bits in a colorized image is at most about 10.8 bits (depending on the color theme used to colorize the original depth image). However, if the original depth image was uint8 then the original depth may be fully recoverable.
The color theme selected must match the theme used when the original depth image was colorized. Note that not all color themes used for colorization of a depth image may be decolorized. The block will issue an error if a color theme is selected that cannot be decolorized.
The polarity of the DSR signal used to enable the LIDAR motor has been inverted for all targets except QCar, since the QCar inverts the signal in hardware to prevent the LIDAR from spinning as soon as the car powers up.
Support for the Leishen MS10 LIDAR and YDLIDAR LIDAR has been added to the Ranging Sensor block. However the Leishen MS10 LIDAR currently does not support Fixed-size interpolated output Output mode.
It is now possible to optionally output version information about the ranging sensor from the block. Version information is useful to know whether the device firmware requires updating.
The Video3D Capture block now has optional outputs for the camera intrinsics and depth scale factor. The camera intrinsics are represented as a 3x3 matrix including the focal length and principal point in pixels.
The depth scale provides the scaling factor between the depth image pixel values and meters. The scale factor will change depending on the output data type. If meters are being output directly then the depth scale factor will be 1.0.
Bug Fixes
Some of the bug fixes in QUARC 2022 are enumerated below. This list is not complete. Only some of the highlights are listed below.
Fixed conversion from the internal Z16 depth format used by some RGBD cameras to MATLAB grayscale uint16 and RGB uint16 on Intel platforms caused by Intel oneAPI functions failing to transpose a uint16 matrix correctly.
The "<<" query button on the Game Controller block parameter dialog has been fixed so that it handles the absence of any game controller correctly. Support for the (Ubuntu) target was also added for the query button.
The Ranging Sensor block now uses Boost mode, if it is available, when the Long range option is selected.
A small bug in the decompression of the RPLIDAR packets was also fixed, which resolves the issue with clustering of adjacent LIDAR scan points in some circumstances.
Plugged a memory leak in the To Host File block when saving compressed images to a video file.
When the display setting for scaling is not 100%, the Video Display block was not presenting the image correctly, resulting in only a portion of the image being shown and the mouse-hover toolbar being invisible. Scaling is now handled correctly.
Copyright ©2024 Quanser Inc. This page was generated 2024-04-16. Submit feedback to Quanser about this page.
Link to this page.