Microscopic Digital Camera Hardware And Its Video Stream Algorithm: Realization And Application

Microscope is a very important optical instrument to research and change the micro world. However, the traditional optical microscopes are not easy to use, for reasons that they will lead to tiredness, the results of observations are subjective, they are not suitable for large-scale unattended tasks and they have few functions. In order to overcome these weak points, several CCD and CMOS digital cameras are designed and implemented to digitize the optical microscopes. Complete digital camera hardware and software solutions are provided. Based on the dedicated digital cameras, a microscope auto-focus system and a multi-focus image fusion system are designed and implemented, which extends the scope of applications for optical microscope.In the design of the digital camera hardware system, a bottom-top idea is adopted. Starting with the selection of image sensor, video data acquisition solution and video stream transfer solution, a concept of multi-configuration for digital camera hardware is proposed. A complete solution with a single set of hardware circuit supporting multiple image sensors is provided based on the dedicated design. Long time running and special purpose tests show that the hardware system of the digital camera has the advantages of low power, low heat, low noise and high stability.In the design of the digital camera software system, the ultimate goals are software reusability and user experience, which are well considered during the design of each soft-ware module. Giving the analysis of the core functions and the extensibility of the system, the software architecture is provided. The whole software system consists of device driver, low-level processing algorithm library, digital camera plugins under multi-media frame-work and software main framework modules. Several theoretical and practical innovations are contained in the low-level video stream processing pipeline algorithms. In terms of soft-ware architecture, program efficiency and user experience, the software system exceeds similar ones at home and abroad. In order to extend the application scope of the digital cameras, and make it a smart optical instrument, two practical systems based on the digital camera are constructed:mi-croscope auto-focus system and microscope multi-focus image fusion system. The bisection +mean square deviation combined with Fibbonacci+Laplacian method is first proposed to meet the speed and accuracy requirements of the auto-focus system. Coordinated with the dedicated circuit and mechanic system, the algorithm is proved to be efficient and stable. Based on the hardware of the microscope auto-focus system, a multi-focus image fusion algorithm based on filter in frequency domain and synthesis in space domain is proposed. The experimental results show that the algorithm performs well in terms of visual evalua-tion and objective quality metrics. This algorithm is easy to use (parameter-self-adaptive) and insensitive to noise and unregistered scenarios.