A large field of view visible cameras image acquisition and processing technology in EAST device

Abstract

A large field of view visible image acquisition and processing system was designed to address the discharge visual needs for estimating the state of plasma operation and the first wall conditions during the plasma discharge of the Experimental Advanced Superconducting Tokamak (EAST) device. The entire system, which consisted of software and hardware, was introduced in detail. Moreover, three sets of the visible camera diagnostic systems with the different field-of-view images acquired by three industrial personal computers (IPCs) were used for forming the large field of view through image stitching. Three IPCs, and data storage and image processing servers are connected by a reflective memory network, which can realize the collection and fast transmission of image data during discharge. In addition, an image stitching algorithm based on block matching scale-invariant feature transformation was adopted. To accelerate the processing of the algorithm, parallelization was achieved by using the Graphics Processing Unit (GPU). Through the mesa test and experimental results, the proposed system provides a feasible way of transmitting and stitching images quickly and fulfills the EAST experimental demands.