| 摘要 |
<p>The number of read pixels N per image sensor chip (A) is expressed by N=n.k (multiples of 16 excluded), where n is the number of image signals per block when a plurality of image signals output from a plurality of photodetectors (2) are level-corrected collectively on a block basis, and k a minimum integer satisfying the condition, 0∫(n.k-η.p)∫n, η being a resolution as the number of read pixels per preset unit length and p being a minimum difference between a plurality of requested read widths expressed in terms of the above unit length used as a reference unit. When a plurality of image sensor chips (A) are used in one-row arrangement, a ratio of an allowance for an image sensor chip (A) to a preset read width is constant.</p> |
