With the rapid development of optical fiber communication technology and micron level high-precision product manufacturing industry,traditional manual and naked eye detection methods not only have large operational errors,but also no longer meet the needs of mass production.In optical fiber communication products manufacturers to improve the product detection accuracy and efficiency,focus on the development of a set of software testing system meets the needs of mass production,and machine vision algorithm development is efficient is the core content of construction of the automatic detection system.In this paper,the target location technology of optical fiber PCB board based on template matching is studied.The main work of this technology is as follows:1)On the basis of traditional SSDA algorithm,a template matching search strategy based on improved SSDA matching algorithm and a Pyramid structure is proposed,which improves operation efficiency and matching efficiency.2)The target recognition system of fiber optic transceiver PCB board is designed,and the software interface of C++ and MFC is developed to realize the functions of automatic transmission,visual recognition and human-computer interaction.Experiments prove that the improved SSDA algorithm and reasonable shape feature points selection can achieve better performance in target matching accuracy and average time consumption of fiber PCB plate.Compared with the traditional template matching algorithm,the recognition performance is improved.The algorithm is also of great significance for the design and development of automatic detection system similar to high precision products. |