Microsoft has begun shipping “thousands” of the final version of the Kinect for Windows v2 sensors to developers, the company announced July 15, several days after it began accepting preorders.
The company first released Kinect for Windows v2 Developer Preview kits, which contained a pre-release version of the hardware, for select developers in November. The move coincided with the launch of the Xbox One, which included a consumer version of the sensor.
Developers will soon be able to put the final Windows-compatible hardware through its paces and begin the process of commercializing apps that rely on the upgraded sensor. Microsoft assured customers that “more sensors will leave the warehouse in coming weeks, as we work to fill orders as quickly as possible.”
Kinect for Windows v2 is a marked improvement over its predecessor, which was based on the popular peripheral for the Xbox 360. Originally developed as a motion and voice control add-on for the game console, hardware hackers quickly pounced on the low-cost device to build 3D mapping prototypes, autonomous robots and a variety of advanced real-world sensing solutions.
Now, Microsoft is betting that the new hardware will help Kinect enthusiasts push boundaries further.
The latest version features upgraded skeletal tracking. Anatomic accuracy is enhanced with the ability to track 25 skeletal joints per person, comparted to 20 joints previously. Kinect for Windows v2 also captures video in full 1080P (1,920 pixels by 1,080 pixels) versus 1,280 pixels by 960 pixels in the original version.
Also improved is the device’s infrared sensing feature, boosting the Kinect’s detection capabilities and enabling developers to more easily handle challenging lighting conditions. Finally, the hardware features an expanded field of view, allowing the camera to effectively capture imagery over a larger scene.
To help developers better exploit those capabilities, Microsoft has also released a public preview of the Kinect for Windows SDK (software development kit) version 2.0. It includes support for the Unity cross-platform development environment and new tools that leverage machine learning to streamline the process of coding gesture-based apps.
In addition, Kinect for Windows SDK 2.0 features multi-app support, granting more than one application access to a single sensor. Face-tracking has been improved with a new a 2,000-point mesh for a more lifelike representation of a person’s facial features. The SDK is free and there are no fees associated with runtime licenses of commercial apps that are developed using the software.
Independent Kinect development has evolved from its hobbyist roots, according to a Microsoft spokesperson. Already, thousands of businesses are leveraging “Kinect for Windows v2 technology to create immersive experiences, including Disney, Pepsi and Genesis Healthcare,” the spokesperson told eWEEK.
Currently, Microsoft is working on the final version of the Kinect for Windows SDK 2.0. When it is finally released sometime later this year, developers will be able to begin submitting compatible apps to the Windows Store.