Design And Realization Of The Automatic Identification And Classification System Of Marine Microalgae

Marine-microalgae occupies an extremely important position in the structure and function ofecosystems, changes in its number and spatial distribution will have an important role inchanging the functions of the entire marine ecosystem and even the global climate. And recentstudies show that marine-microalgae also bears a great value of commercial, production and life,so accurately identifying the useful microalgae is becoming the problem’s key.In recent years, using digital image processing technology to research marine microalgae’sautomatic identification and classification develop fast. In this paper, we under the premise ofdoes not distinguish the types of microalgae, achieved the automatic image recognition andclassification system with the microscopic image acquisition technology, digital imageprocessing, pattern recognition, database storage technology and VC++graphical user interfaceimplementation technology and other related marine microalgae technology. The paper firstlyintroduced the hardware configuration and acquisition process of the marine microalgaemicroscopic image; then introduced the corresponding pretreatment to the randomly collectedmicroalgae mixed sample images, including converting color image to gray, image enhancement,and binary image analyzing, also verified by experiments or theoretical basis selected theoptimum method; then through corresponding algorithm extracted the rectangular shapecharacteristic value of the target microalgae and the texture features of Energy, Entropy,InertiaQuadrature, LocalCalm, Correlation; Finally, achieved the automatic identification andclassification system with biomimetic pattern recognition algorithm based on Euclidean distance.The experiment proved that the system is not only able to identify different types of algae, butalso can identify the different sizes of the same type, also with high recognition accuracy, highspeed,has feasibility.