New version of Kinect for windows SDK1.6.0 released

New version of Kinect forwindows SDK1.6.0 released

Today () Microsoft released a new Kinect SDK, which is now available for windows through official channels.

Official introduction to SDK1.6.0 website

Http://msdn.microsoft.com/en-us/library/jj663803#SDK_1pt6_M2

Http://www.microsoft.com/en-us/kinectforwindows/develop/new.aspx

Blog master introduction:

Http://www.cnblogs.com/yangecnu/archive/2012/10/09/New-features-in-Kinect-for-Windows-SDK1_6.html

Added sensor data operations

1. Exposure of the API's infrared light as a new color image format

2. Extended sensing distance (more than 4 meters). As the distance increases, the imaging quality will decrease. This is restricted by Microsoft software!

3. The camera settings are further optimized.

4. Support for a new Bayer RAW color format, that is, the most primitive data. Before that, Microsoft converted us into rgb, yuv, and other formats (the encapsulation is good ).

5. Exposure of the sensor accelerator in the API

6. Support for German language recognition packages

7. Several new APIs supporting color, depth, and coordinate architecture conversion. Both the single point and the entire image are converted. I haven't tried it yet, but I have seen many functions starting with map in the C ++ function, which is really awesome. There are also some functions that map color images to deep images, which have never been before.

8. Supports Windows running on virtual machines, including Microsoft's Hyper-V team, VMWare, and Parallels.

Support win8 and visual studio 2012.

The API for color conversion is convenient: White Balance, contrast, tone, saturation, etc.

At the same time, Kinect Studio1.6.0 supports the storage and reading of infrared, raw Bayer, extended depth data, and accelerometer features.

A funny message about the release of kinect in China: Great, anothertechnology the Chinese can steal and reply back to us.

After I installed the new SDK, I restarted my computer and started the previously developed Kinect Project (based on SDK1.5.1), re-compiled and run it. There is no problem!

Next, let's take a look at the new example!

1. BasicInteractions wpf c # Sample

Displays a full screen human-machine interaction application. Both speech recognition and gesture recognition are supported. This operation is the same as that of the xbox application. Microsoft provides a standard for developers to develop applications in imitation. People who develop human-computer interaction must carefully read this example. It's C #, OKay ~

2. Infrared Basics-D2D C ++ Sample

The original infrared data is displayed! This is very different from the depth image we see, the previous is processed (calculate the time difference between the near infrared reflection, get the distance, give a URL http://blogs.howstuffworks.com/2010/11/05/how-microsoft-kinect-works-an-amazing-use-of-infrared-light/ that needs to flip the wall ), the current data is actually raw. This is consistent with the previously seen infrared image. In this case, researchers can study better reflection algorithms to make the obtained deep data more accurate than Microsoft's! This is open source.

Infrared data can be collected at the same time as the depth data, rather than the color image.

3. Kinect Explorer

Almost all newly added items can be seen in this example, that is, the newly added API makes development easier. Originally, we used OpenCV and other tools to process images. Now the SDK comes with it!

What's interesting is that when we tilt the kinect, the value changes! I still don't know the specific reason. This development robot-related application will be used, and will not be very useful for applications with fixed locations.

In general, Microsoft has published more bottom-layer data for kinect, allowing researchers to study the bottom-layer data. Then, Microsoft provides more practical functions and examples to help developers develop applications. Both the underlying layer and the upper layer have been upgraded.