This work focuses on reducing false execution of brain-computer interface(BCI) based Soft Robotic Glove by considering visual information received by the first-person-view camera equipped by the user. The proposed method intends to seek to lower the false positive rate while providing an intuitive interface by allowing the glove to execute depending on motor imagery(MI) only when the hand is in sight. When the hand is out of sight, no electroencephalogram(EEG) information is given to operate the glove. Two sessions of online soft robotic glove control were conducted on six participants, one session each for BCI-based and BCI/Vision-based glove control. The result showed that using visual information with BCI helped the participants remain rested than they did with BCI-based soft robotic glove. However, additional experiments from different participants are necessary to ensure the effect that using visual information could have on grasping and releasing action of BCI-based soft robotic glove as well.