An Exploration of Bloodstain Aging Analysis Methods Based on Clustering Algorithms

基于聚类算法对血迹陈旧度研究方法的探究

This study employs computer technology to analyze RGB values during the process of bloodstain change, primarily using the K-means clustering algorithm, image segmentation, and polynomial regression analysis. It explores a novel approach for determining the age of bloodstains. Additionally, Pearson correlation coefficients were applied to assess the effects of temperature and humidity on the morphological changes of bloodstains. The integration of computer technology into bloodstain pattern analysis offers an objective method for estimating bloodstain age, reducing the influence of human factors on results and enhancing case investigation and resolution efficiency. Blind test results indicate that the method achieves an accuracy rate of 85.7% within a one-hour margin: a 10-minute deviation was found in 17.96% of cases, a 20-minute deviation in 37.5%, a 30-minute deviation in 55.4%, and a 60-minute deviation in 85.7%. Overall, this method demonstrates a promising accuracy rate for bloodstain age estimation.

研究基于计算机技术对血迹颜色R、G、B 随时间的变化规律进行分析，主要方法包括Kmeans聚类算法、像素聚类，图像分割、多元多项式回归分析，探索出可用于血迹陈旧度分析的新方法，同时利用皮尔逊相关系数对18 组环境条件开展的实验结果进行分析，探究本实验中，温度、湿度对血迹形态变化的影响作用。将计算机相关技术应用于血迹学研究领域可提供一种客观的方法分析血迹陈旧度，减少人为主观因素对分析结果的影响，更可以提高侦查、破案效率。本研究中所提方法用于血迹陈旧度推算的盲测结果准确率为推算时间与实际时间相差10min 占比17.96%;相差20min 占比37.5%;相差30min占比55.4%;相差60min 占比85.7%。综合来看，应用该方法推测血迹准确度，推测时间误差在1h 内的准确率可达85.7%。