A Study on the Automatic Arm Vessel Injection Point Judgment by Applying Machine Vision Techniques

• Main Authors: YUN-SHENG YEH, 葉昀昇
• Other Authors: CHENG-HO CHEN
• Format: Others
• Language: zh-TW
• Published: 2017
• Online Access: http://ndltd.ncl.edu.tw/handle/cj5gbw

碩士 === 國立勤益科技大學 === 機械工程系 === 105 === Machine vision technology uses a computer, video capture card along with industrial cameras, lens and lighting and other optical equipment, to find out the surface characteristics, geometry or position of the object under test through software analysis. It replaces the need to manually examine the object. So far, many industries have begun to the incorporate it with mechanical arms and other drive units. It is also an indispensable technology in industrial 4.0 and machine learning. The technology has been used in a more extensive range of applications with innovations in the hardware equipment and manufacturing methods. Precision and stability have also been greatly upgraded. In the medical industry, there are also many applications of the image processing technology, for example, to assist physicians, nurses and other medical practitioners to determine the physical conditions of the patient, make the appropriate diagnosis, and also reduce the fatigue or other human factors causing misdiagnosis. In this study, we discuss how to use machine vision technology of to assist the blood collection and drug injection by robotic arms. In this thesis, CMOS industrial cameras are used with a infrared IR-810nm light source. Camera lens are installed with a double blood vessel images taken through 780 ± 10nm filters, The cameras are calibrated beforehand using the machine vision software HALCON. The images are then processed through image binarization and morphological algorithms. After image processing, the best needle position, needle depth, puncture depth and angle are determined.