********************************* *********************************
(Pattern Recognition) 1 1. CCD 2. 3. 4. 1
ABSTRACT KeywordsMachine Vision, Real Time Inspection, Image Processing The purpose of this research is to develop an in-process machine vision system using CCD. The system will be designed to inspect the marker defect of aspect on package. The method of image processing is planned to utilize the Pattern Recognition. The lighting system is intended to adopt the infrared LED for the merit of good efficiency, low frequency band, and low noise. Also the lights are set on the top of object for obtaining clear picture. Microprocessor will be used to control the transmission system with hand shaking with the personal computer. The infrared sensors also are planed to control the position of the object. The inspecting time from position, capturing picture, image processing, to get the result will be within 1 second. This predicted achievement of this system includes: 1. Set up the accessory control technique of Machine vision system for dynamic inspection, including the strobe light and the control of CCD. 2. Study and program an efficient and fast running software for fulfilling the real time Inspection on marker defect. 3. Design a transportation mechanism system by using Microprocessor. 4. Develop man/machine interface to combine afore-described systems into a mechanic prototype system. 2
3 ; [1] [2] 1 ;2 3 ;4 5 6 7 [3][4][5]
CCD. 4
CCD I. PICOLO 640X480 256 CCD Camera II. Pentium 550MHz 8255 8952 III. [2] IV. 8952 8255 [7][8][9] 5
(Pattern Recognition) ( ) (1) (2) r r r 100% r 0 (1) (2) r = i= N i= 1 I i M i 2 2 2 [ N I ( I ) ] N M ( M ) 2 [ ] (1) N IM I M r = (2) M I Logo Logo Logo Logo Template Blob Blobs Logo [15] Logo Region Growing 6
7
[16][17] 2 (Block Matching Criterion) 2 90% 15% Blob Pixel Blob Pixel 8
No No Yes Yes Yes No 9
Windows Visual Basic 6.0 98% 2 IC 88 90 2 1 60 59 (88 + 59) 100% = 98% (90 + 60) 10
.. 11
.. 12
. 13
1. 116 176-184 1992 2. 123 267-279 1993 3.,, 1998. 4., 1993,, 5. (86),, 6.,, 1993, Bit Mapped, 7. (86)8051/8052 8. (84), IBM PC XT/AT, 9. (87), EM 78x56, 10. (87),, 11. (80),, 12. Roland T. Chin, Survey: Automatic Visual Inspection: 1981 to 1987, COMPUTER VISION GRAPHICS AND IMAGE PROCESSING 41, 346-381, 1998. 13. Baptista, G and Kulkarni, K.M. A High Accuracy Algorithm for Recognition of Handwritten Numerals Pattern Recognition. Vol.21, No. 4, pp.340-344, 1992. 14. Pavlidis T., Kahan, S. and Baird, H.S., On the Recognition of Printed Characters of Any Font and Size, IEEE Transactions on Pattern Analysis and Machine Intelligence, Volume PSMI-9, No.2, pp.274-287, 1987. 15. W.H.Tsai, Moment-Preserving thresholding : a new approach, Computer Vision,Graphics and Image Processing, 29, 1985, pp.377-393 16. R.Li, B. Aeng and M. L. Liou, A new three-step search algorithm for block motion estimation, IEEE Transactions on Circuits System Video Technology, vol. 4, pp.438-442, August 1994. 17. J.Lu and M. L. Lious, A simple and efficient search algorithm for block-matching motion estimation, IEEE Transactions on Circuits System Video Technology, vol. 7, pp.429-433, April 1997. 14