In:
Journal of Applied Physics, AIP Publishing, Vol. 103, No. 7 ( 2008-04-01)
Abstract:
This study proposes a new miniaturized autofocusing mechanism and an intelligent autofocusing algorithm with digital images as feedbacks. An actuation scheme for autofocusing is fostered via (1) capturing image by a complementary metal-oxide semiconductor sensor array, (2) evaluating the clearness of images by the well-known method of discrete cosine transformation (DCT), and (3) actuating the lens in the designed mechanism to the best-focusing position estimated by a neural network (NN), assisted by a genetic algorithm (GA).
Type of Medium:
Online Resource
ISSN:
0021-8979
,
1089-7550
Language:
English
Publisher:
AIP Publishing
Publication Date:
2008
detail.hit.zdb_id:
220641-9
detail.hit.zdb_id:
3112-4
detail.hit.zdb_id:
1476463-5
Permalink