Conversion method for obtaining data for quickly identifying injured individual based on pelvis CT picture

The invention discloses a conversion method for obtaining data for quickly identifying injured individuals based on a pelvis CT picture. The conversion method comprises the following steps: 1) performing CT scanning on a plurality of individuals; 2) collecting pelvis CT pictures of the individuals; and 3) performing feature extraction on the pelvis CT picture by using a convolutional neural network feature extractor to obtain feature data, and converting the feature data by using a conversion equation to obtain converted comprehensive feature information data. According to the method, the features can be extracted from the pelvis CT picture simply, conveniently and rapidly with low cost, the features are converted to obtain converted comprehensive feature information data, and the comprehensive feature information data is further used for a rapid identification model of individual injury. 本发明公开了一种基于骨盆CT图片获得用于快速识别受伤个体的数据的转化方法，包括以下步骤：1)对若干个体进行CT扫描；2)收集这些个体的骨盆CT图片；3)利用卷积神经网络特征提取器对骨盆CT图片进行特征提取，得到特征数据，并利用转化方程对这些特征