InnovationQ will be updated on Sunday, May 20, from 10am - noon ET. You may experience brief service interruptions during that time.
Browse Prior Art Database

Stabilising mobile projection by culling boilerplate

IP.com Disclosure Number: IPCOM000244585D
Publication Date: 2015-Dec-23
Document File: 5 page(s) / 72K

Publishing Venue

The IP.com Prior Art Database


As projectors become more pervasive there has been a large push to minimise the size of the hardware required. As the technology shrinks further it will be possible to create mobile projectors that are part of phones, or trivial to carry with a phone (e.g. in a case). A simple USB connection is enough for android devices to power and display to a projector. This will present a new realm of problems - namely, the stabilisation of picture from the projector when on a mobile device.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 01 of 5

Stabilising mobile projection by culling boilerplate

This article seeks to solve the problem of projections moving as a phone moves by taking into account the motion of the phone and surface. If the phone moves the projection will be moved into 'boilerplate' sections of projection - those which do not affect what is being displayed. This will keep the projection in the same physical space, stabilising it.

Phone projecting onto a surface

Problem: As the phone moves the projection moves with it

Solution: Track phone movement to stabilise the projection against phone movement

There are two core sections:

Tracking the phone's location compared to the projection space
Moving the 'interesting' part of the projection into the 'boilerplate' part of the projection
Phone tracking:

Page 02 of 5

    We initially track the projection space when the projection is started. This gives us the area where the projection will be stabilised to. Once we have this initial projection space we then track the phone in space or image features in relation to this space. The first relies on a 6-axis gyroscope in the phone, which should be

standard in future. It is possible to take the velocity & acceleration of the phone and convert it to a position in 3d space. Tracking this position as it moves from the initial projection location gives an offset which will feed into the boilerplate culling. The second way to track this is via image features using the on board camera. This can be done exactly the same as markerless augmented reality - image features are picked out of the scene and tracked from the camera's point of view. As it moves in pixels, this is sent to the boilerplate culling algorithm.

Boilerplate culling:


Page 03 of 5

     There are two distinctions that are made - 'boilerplate' sections and 'interesting' sections. Taking t...