查看文章

Ultrasound imaging is one of the most widely used biomedical imaging modality for diagnostic purposes. Ultrasound image segmentation is an important step in order to achieve efficient qualitative measures, such as location of relevant structures, and quantitative measures, such as area of the structures for further analysis. However, segmentation of ultrasound images is considered a challenging task owing to its poor quality. Moreover, segmentation using global image statistics results in unsatisfactory results, making it difficult to use for further processing. In this study, we formulate a scheme to achieve successful segmentation of real time ultrasound images. The proposed method makes use of local statistics of the image in order to perform efficient segmentation. It considers histogram-based thresholding by making use of recursive minimum cross entropy on statistical properties (image integral and image …