Target Images

Although the ImageTrainer program will accept any PNG or JPEG image as input, there are a number of desirable image characteristics that will improve the tracking performance of your Popcode scene.

Amount of texture: Popcode uses the texture of the image in order to work out the position of the image in the camera view. A single block of colour or subtle gradients will therefore not work well.

Unique texture: If the image contains lots of repeated structures it is difficult for Popcode to work out the position of the camera. For example, imagine a chessboard with the camera positioned such that only 4 squares are visible. There would be many possible camera positions that would produce an image which looked the same, and therefore Popcode would be likely to pick one at random, which would mean your content would be likely to appear in the wrong position.

Spread out texture: Ideally images should have texture across their entire area, or at least across the area where you want the augmented content to appear. Users may want to move closer to your target image to get a better look at a particular part of the scene, but if that part is in a plain area of the image Popcode is likely to be unable to work out the camera position.

Size: The ImageTrainer program will offer to resize your input images for you if they are over 640 pixels in width or height. We recommend this as the maximum size for your target images. It is possible to train images larger than this, but this should only be necessary in rare cases (such as wanting to operate very close to to large printed targets). Training will potentially take a very long time with large images. For applications with small targets (such as business cards) you may find resizing your input image to a smaller size before using the ImageTrainer program will give better results.

In summary, the best target images for Popcode are those that have a good spread of unique texture. In practice many images work well; the draft training mode will give a rough idea of whether a particular image is suitable.