The Design And Implementation Of Urine Test Strip Quantitative Detection System

The urine test strip quantitative detection system is a portable medical detection system which can be used at home.This system has high requirements for optical imaging,device design,software design and image processing.Compared with the traditional urine detection methods,the device's portability and fully automated detection method are helpful for popularization,and it proposed and solved the quantitative problem in the urine detection process for the first time,which makes the test results more accurate and objective.The design and implementation of the urine test strip quantitative detection system mainly includes system devices,algorithms and software.The device is designed by SolidWorks software tools,the algorithm is implemented by C++ advanced programming language,and the software system is implemented by Java advanced programming language.It is divided into a server running on Linux operating system and a client running on Android operating system.On this basis,the main research content and research results of this topic are as follows:First of all,this paper designs a portable quantitative urine testing device for the shortcomings of traditional urine testing instruments such as bulky,expensive,and inability to control the dropping volume of urine samples.The volume of the device is only about 10% of traditional urine testing equipment;the manufacturing cost is only about 5% of traditional urine testing equipment.And the operation mode of the institution is simple,and the real urine quantitative test can be completed without professional medical knowledge.Secondly,this paper proposes a color vector similarity evaluation method based on image subtraction,which comprehensively considers the similarity of image colors from multiple aspects.It is more objective and effective than the traditional color similarity evaluation method.In addition,this paper proposes a fully automatic image color classification algorithm combining BP neural network and SVM in view of the shortcomings of traditional urine detection methods such as slow detection speed and the inability to guarantee the accuracy of manual detection results.This algorithm combines the advantages of the image color classification algorithm based on BP neural network and the image color classification algorithm based on SVM.The accuracy of the image color classification is improved to more than 94% through the trained result model.The image classification process is fully automated,and the efficiency of single detection is about 20 times faster than that of traditional human detection methods.Then,in view of the shortcoming that the traditional urine detection method can not control the reaction time of urine samples and urine analysis test strips,a video is used for the first time to establish a color change model during the reaction process of urine analysis test strips to determine the best response time of the test strips.Combined with the timing module in the device in this article,the measurement accuracy of this system has reached a new level.Finally,this article designs the main flow of urine testing.According to the results of 20,000 test requests,the average processing speed of the system can reach 800 per second,which can be used by 200 people online at the same time.This shows that the system can run at high speed and stability,meeting the current performance requirements.