The Design Of Blood Lancet Real-time Image Detection Algorithm

With the development of computer technology, the digital image processing tech-nology has been widely used in the industrial field, especially in product testing significantlyefective. In the modern production process, the detection system based on digital image pro-cessing technology has gradually replaced manual inspection, and continuously improves thedegree of automation of industrial production.Currently in the production process of the blood lancets, although automatically takeingand puting lancets have been achieved, but now the way of detecting blood lancets placed instate is artifical. People is vulnerable to the impact of the labor intensity,which causes theemergence of the missed and false detection problems,and can not guarantee the quality ofthe product.At the same time,it has high production costs and low production efciency.Tothis end,this paper combines with actual production needs, based on the the blood lancetsequipment image features,and put forward the real-time detection system based on digi-tal image processing technology. At the beginning, the paper introduces the blood lancetsproduction, and so raises the design ideas of the lancets real-time image detection system.To the algorithms in the the blood lancets real-time image acquisition process, includingimage preprocessing and image segmentation et cetera, it carries out related presentations.Then through learning cascade fast object detection algorithm based on Adaboost learningalgorithm, including Adaboost learning algorithm, rectangular features and integral image,based on the advantage of this algorithm based on auto-learning feature,quickly and ef-ciently achieving the target detection,the proposed system completes target detection. Thenintroduce the design process and the realization method of the paper in great detail. In theprocess of designing the system, herein proposed Unique lancets real-time image acquisitionmethods to accurately obtain the target image.Based on the classic cascade rapid target de-tection algorithm based on Adaboost learning algorithm, to blood lancets image features,weget efective extraction and analysis. Finally, for the system we make experimental analysisand summary.This article uses the Microsoft Visual Studio2008Chinese the version, OpenCV2.0de-velopment software system.It uses modular programming methods. It is convenient for pro-gramming,operating and processing, and makes the operation interface concise and friendly.In this paper,to the system designed,we carry out a series of experiments and results analysis.Based on the final results show that the system has fast detection speed, good real-time per-formance and high overall detection accuracy rate.This provides a reliable technical supportfor the actual production automation.