Activity testing model for automatic correction of hand pointing

•We propose an activity testing model to assess automatic correction of hand pointing.•The mechanism of automatic correction of hand pointing provided the theory basis.•Two types of hand pointings were calculated and tested.•The score from the proposed scoring system can provide sufficient estimation.•Our testing model was effective and it could be applied to neurorehabilitation.

In this paper, an activity testing model was proposed to detect and assess automatic correction of hand pointing. The average recognition rate for automatic corrections of hand pointings was 98.2% using the acceleration data. Moreover, a score was calculated using the activity data of successful recognition and it provided sufficient estimation for the performance level of automatic correction. Experimental results showed that our model was effective and it could be applied to neurorehabilitation.