Computer generated spatial filters for large object formats

Abstract A cheap computer generated hologram (CGH) contains 32 × 32 sampling cells, typically. When using a cheap CGH as spatial filter, the input object is also restricted to 32 × 32 sampling or resolution cells. However, practical objects are much larger. Sufficiently large filters would be too expensive in terms of computing and plotting time. We present a new method for producing CGH spatial filters. We were able to produce a 160 × 160 cell CGH filter with computing and plotting labour for only 32 × 32 sampling cells. We used that spatial filter as matched filter for pattern detection.