FAQ

Common

We ONLY support using 64-bit Windows 7/8/10 and Linux Ubuntu 16.04 platforms for your developments with the SDK.

*If you really want to use SDKon other platforms (eg. NVIDIA TK1 / 64-bit Ubuntu 14.04 / etc.), please email us for further customization.

The LeadSense camera is based on passive stereo vision. Unlike active depth sensors, the LeadSense camera replicates the way human vision works. For example, if you can see something, the LeadSense camera will most probably see.

If the environment is low-lignt or texture-less, the depth accuracy will be decreased due to the weak information available in the image, or may not be able to calculate depth.

Many manufacturers currently distinguish Type-A USB 3.0 ports with blue inserts, meanwhile USB 2.0 ports will use black inserts.
And sometimes, the USB SuperSpeed logo, or its “SS” abbreviation, may be used to identify USB 3.0 ports.

Camera is calibrated before putting on sale. If you find that the distance result is not correct, you can use CalibrationTool provided in the SDK.
IMU calibration is related to the environment. It is strongly recommend to re-calibrate the IMU when you try to use it in a new environment.

The usage of camera & IMU calibration tools provided in the SDK, please check Tools section in the document.

Also you could take a look at Video tutorials.

Since EvoBinoSDK 0.8.0, we provide one more interface evo::bino::StereoCamera for no CUDA environment.

You can using evo::bino::StereoCamera to get images and IMU data just as using evo::bino::DepthCamera.

Since EvoBinoSDK 1.0.0, you can also get depth information when using evo::bino::StereoCamera. evo::bino::StereoCamera only uses CPU to do the depth calculation, and speed may be slower than using evo::bino::DepthCamera.

Note: If you just want to record image sequences(.evo) or just want get images with no depth information, please use evo::bino::StereoCamera. Because frame lost may be occurred due to the upload/download speed of GPU when using evo::bino::DepthCamera.

Troubleshooting

Check MD5 on SDK’s hash page, make sure that the installing package is correctly downloaded.

When using camera with high frame rate (eg. 120fps), the amount of data transmission is very large. If the power supply from the PC’s USB controller is unstable or the USB controller doesn’t have available bandwidth, image data may be corrupt.

Recommended solution:

1. Try to use a USB 3.0 hub to solve the problem of unstable power supply.

2. Try to use a PC with a better performance USB controller.

3. Try to select a lower frame rate.

The VC++2013 run-time libraries are required by the tools in SDK.

This error means that VC++2013 run-time libraries is missing.

Try to download Visual C++ Redistributable Packages for Visual Studio 2013 from Microsoft official site.
After the installation, this problem should be fixed.

The NVIDIA CUDA toolkit is required by the SDK to run depth computation.
This error means that NVIDIA CUDA toolkit is missing or that the wrong version is installed.

Determine the needed CUDA version from the last characters of the filename.
(For example, _80 stands for CUDA 8.0, _92 stands for CUDA 9.2)

Try to download CUDA from NVIDIA official site.
After the installation, this problem should be fixed.

LeadSense will be the default audio input device after it has be connected. Some types of LeadSense has embedded microphone, so you can use it directly as your microphone.

*If you are using LeadSense without microphone, there will be no audio input from LeadSense. But default audio input device will still be changed. So please manually change the audio input device to the previous one after you connect LeadSense.