| 摘要 |
<p>A computerised methodology for reducing swath scanner swath boundary discontinuities is provided by using a predetermined overlap of pixel row during each sequential scan. Pixel data from a first scan (400-401) of the overlapped pixel rows is compared to the pixel data from a second scan (405, 409, 411, 415, 417) of the overlapped pixel rows, identical pixels of the overlapped pixel rows ideally having the same representative scan data. Errors are compensated by weighting representative scan data for each pixel row, linearly decreasing the weighting in proportion to the proximity to the swath boundary (413). Weighted data is substituted for true data depending upon comparison factors (403, 407). In one embodiment, the weighting factor is an increase or decrease in the first scan data relative to the percentage change in white level intensity captured in the second scan of the overlapped pixel rows. In another embodiment, the weighting factor is a linear shifting of the proportion of the pixel data from the first scan and the pixel data from the second scan of the overlapped pixel rows used as the scanned image data such that the proportion of the first scan decreases as the proportion of the second scan increases, thereby smoothing differentials in data representative of individual pixel rows at swath boundaries. In another embodiment, scanner signal-to-noise ratio is improved by using a weighting factor in accordance with the Central Limit Theorem. <IMAGE></p> |
