摘要

An automatic tonsillitis monitoring and detection system aims for personal use and requires mobility, a compact size, and a light weight with reliable functionality. This paper proposes the design and implementation of automatic tonsillitis monitoring and detection system. In this system, a pipeline concept is utilized for data transmission among processes, and parallel processing is employed for feature extraction by dividing an image into an appropriate number of blocks for processing. The performance of the proposed method was evaluated by experiments using prototype software and hardware systems of our design, and the results show accuracies of 96.4% and 91.8%, respectively, for 159 tonsillitis samples.

  • 出版日期2017