The Design Of The Saybolt Chromometer Based On Image Processing

At present, China’s national standard and international general oil color detection device is Luo Weipeng colorimeter or Saybolt Chromometer, but the two colorimeter use visual method, there is a big subjective error, therefore taking a more objective technical means to measure oil color becomes necessary.People identify the color through the eyes, brain and life experience producing a visual effect to light, while the computer is the color digital, symbolic, through studying and calculating the distribution of color digital to decide the classification of color. This article uses the camera to replace original viewer of the Saybolt Chromometer, collecting image transfer to PC for processing, at the same time, according to the particularity of oil colors, using color histogram as color feature of the images in the process of research. Using the SVM support vector machine processing image combined with hardware closed-loop feedback mechanism of intelligent measuring oil color.The main work is as follows:(1)Reading a lot of references at home and abroad, introducing the research status, application fields, of color recognition, proposing the design of Saybolt Chromometer based on image processing; In the image processing, the Opencv combined with SVM support vector machine realizes the intelligent colorimetric;(2)Using SVM multi-classification classifier identify the collected images, thus achieving colorimetric; At the same time, through the related hardware setting up closed-loop feedback control input / output relation, so there is a remarkable improvement in the color classification control.According to the classification principle of SVM support vector machine, operation process of the Saybolt Chromometer, and the principle of image processing designed the Saybolt Chromometer based on image processing, which can realize automatic suction and discharge of oil, and automatically determine oil color, having a certain application value.