Automatic Radiographic Position Recognition from Image Frequency and Intensity.

摘要: Purpose. With the development of digital X-ray imaging and processing methods, categorization analysis massive radiographic images need to be automatically finished. What is crucial in this automatic retrieval recognition position. To address these concerns, we developed an method identify a patient’s position body region using only frequency curve classification gray matching. Methods. Our new combined with image The was determined from similarity amplitude classification. performed by matching whole-body phantom prior knowledge templates. stitched radiological different parts. Results. proposed can retrieve recognize intensity information. It replaces 2D 1D classification, higher speed accuracy up 93.78%. Conclusion. able outperform image’s limited time cost simple algorithm. information radiography make quicker more accurate.