| 摘要 |
Technologies are generally described for aligning objects in augmented reality. In some examples, a processor may be adapted to receive detected image data and virtual object data. In some examples, the processor may further be adapted to generate and apply weights to log-likelihood functions at intensity and feature levels based on the virtual object data and detected image data. In some examples, the processor may further be adapted to add the weighted log-likelihood function at intensity level to the weighted log-likelihood function at feature level to produce a cost function. In some examples, the processor may further be adapted to determine transformation parameters based on the cost function that may be used to align the detected image data with virtual object data.
|
