基于快速分水岭分割的火炮内膛损伤检测研究

Research of Damage Detection for Artillery Bore Based on Fast Watershed Segmentation

小口径火炮炮膛表面质量缺陷的检测使用光学内窥仪采集图像后对图像进行分析,这种方法在狭小空间且构件多的火炮炮膛检测中,无法克服亮度不均匀的情况,造成检测结果的不准确,一直依赖传统光学窥膛仪用肉眼定性观测,不能定量检测,准确度差,效率低下,而且检测结果受限。提出一种基于快速分水岭的火炮内膛损伤检测研究方法,使用工业CCD摄像机获得火炮炮膛的清晰图像后,使用快速分水岭的图像分割方法,保护图像的有效边界,避免过度分割。实验证明该方法具有高准确率的火炮内膛检测效,具有很强的实用性。

Small artillery bore surface quality defect detection using optical endoscopic instrument collection after image toimage analysis, this method in the narrow space and member of artillery bore detection, can＇t overcome brightness unevensituation, cause the results of detection not accurate,Always rely on the traditional optical sight bore gauge with the nakedeye qualitative observation, not quantitative detection and the accuracy is poor, the efficiency is low, and the test results isbad.We proposed based on fast watershed in the chamber of the artillery damage detection methods, the use of industrialCCD camera for artillery bore clear image, using fast watershed image segmentation method, protect the image of the effi-cient frontier and avoid excessive segmentation. Results show that this method has high accuracy in the artillery bore detec-tion effect, have very strong practicability.