| 摘要 |
<p>In automatic photographic or electronic camera apparatus, sounds from a predetermined person are preferentially received or identified to provide an audio signal which is then used to produce a saliency signal 11 indicative of the person's response to circumstances to control the camera 1 or an image signal 2, 6 produced therefrom. The sounds may be detected by a sound conduction microphone 4, e.g. on the head or throat of the person, or by voice identification circuitry used with a less specific microphone not necessarily mounted on the person. The person may or may not be wearing the camera. Control may be in real time, or the sound or saliency signal may be recorded 3 together with an image signal for subsequent control 7 of the latter. The audio signal may be categorised by waveform analysis, e.g. for the recognition of sounds such as speech word recognition and/or prosody or non-speech vocal sounds, and/or non-vocal sounds signal amplitude may also be taken into account.</p> |
