No, at this time, the Vuforia API does not expose an API to adjust the "tolerance" (e.g. number of detected features or so...) on the recognition of a target.

Maybe something you could post in the Wish List:

https://developer.vuforia.com/forum/general-discussion/wish-list

Hi allessandro,

Thanks for the reply.The problem  is the book is already out in the market.We cant edit the graphics.My only hope is to manage tolrence level.is there any way in the script where i could control the tolrence level of the image recognition.for eg i would like to play the animation only if it recognizes 90 % or 95 of the details or points ?????? 

Hi allessandro,

Thanks for the reply.The problem  is the book is already out in the market.We cant edit the graphics.My only hope is to manage tolrence level.is there any way in the script where i could control the tolrence level of the image recognition.for eg i would like to play the animation only if it recognizes 90 % or 95 of the details or points ?????? 

Vuforia can detect / track a target with some level of tolerance on the total set of features, e.g. even if only a part of the total features are visible in the camera view; this is done to generally ensure good tracking quality and detection capabilities even if the target is not completely (100%) visible (for instance if it is partially occluded by an obstacle or partially out of view). 

When targets are similar and a very significant part of the target is identical (like the "DOG" part in your case), this can lead to the issue you observed. One way to avoid this would be to differentiate the part of the image that contains the DOG  by editing / adding some small details.