| 摘要 |
PURPOSE: Video motion detection is provided to accurately and efficiently identify motion under environments having excessive illumination changes. CONSTITUTION: A plurality of images of a video sequence is combined on image volume(S1). The image volume is sliced along with one (X,t) plane or (Y,t) plane(S2). Log conversion is performed for the image volume(S3). The image volume is filtered by a time-space filter(S4). A threshold function is applied to the filtered image volume(S5). Complete motion detection of the video sequence is performed by a change of a slicing direction(S6). |
