Title | ||
---|---|---|
Automated Construction Of Image Processing Procedure Based On Misclassification Condition |
Abstract | ||
---|---|---|
This paper proposes a new expert system IMPRESS-Pro that automatically constructs an image processing procedure to decide whether there exists a specific figure in an image or not based on the requirement of misclassification rates given by a user. This system has a sequence of several local processes whose algorithms and parameters are unfixed and decides them sequentially in accordance with condition for each local process performance estimated from the requirement on misclassification rates. This paper also shows experimental results of applying this system to procedure generation problems for extracting defective parts on the surface of integrated circuit packages with promising results. |
Year | DOI | Venue |
---|---|---|
2000 | 10.1109/ICPR.2000.906104 | 15TH INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION, VOL 2, PROCEEDINGS: PATTERN RECOGNITION AND NEURAL NETWORKS |
Keywords | Field | DocType |
agricultural engineering,integrated circuit,feature extraction,quality control,expert systems,image classification,expert system,construction industry,image processing,integrated circuit packaging | Data mining,Pattern recognition,Existential quantification,Computer science,Expert system,Image processing,Feature extraction,Artificial intelligence,Contextual image classification,Integrated circuit | Conference |
ISSN | Citations | PageRank |
1051-4651 | 1 | 0.36 |
References | Authors | |
0 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Toshihiro Hamada | 1 | 1 | 1.72 |
Akinobu Shimizu | 2 | 1 | 0.36 |
Jun-ichi Hasegawa | 3 | 221 | 61.17 |
Jun-ichiro Toriwaki | 4 | 578 | 136.04 |