基于改进Camshift的竹材加工目标检测跟踪算法研究  

Research of Improved Camshift Algorithm in Object Detection and Tracking for Bamboo Processing

作　　者：任洪娥[1] 马晓珍[1] 董本志[1] 王海丰[1] 

机构地区：[1]东北林业大学信息与计算机工程学院,黑龙江哈尔滨150040

出　　处：《安徽农业科学》2014年第24期8455-8458,共4页Journal of Anhui Agricultural Sciences

基　　金：国家林业公益性行业科研专项(201204715);东北林业大学大学生科研训练计划项目(KY2013010)

摘　　要：将计算机视觉技术应用到数控剖竹机运动加工目标的检测和跟踪中,提出一种基于改进Camshift算法的适合竹材加工运动目标检测和跟踪算法。针对竹材检测、跟踪过程中的干扰因素,通过图像的色度值来代替背景图像的亮度值,来减少阴影干扰,采用背景差分法与帧间差分法相结合的目标检测方法,改进Camshift算法,利用HSV图的H分量均值和每一帧H分量均值的差值结果来进行H分量均值更新,以克服光照影响,并利用Kalman滤波实现对下一帧竹材所在位置进行预测,预测结果用于修正Camshift算法的跟踪结果。结果表明,改进的算法能够对运动竹材目标进行实时跟踪,算法高效、准确。Applying the computer vision into detecting and tracking of processing objective of Sectional Bamboo Machine,a new tracking algorithm based on improved Camshift algorithm was put forward,which is suitable for bamboo processing.This paper mainly includes the object detecting algorithm based on background subtraction and frame difference and the object tracking algorithm based on Camshift and Kalman filter,the bamboo position in next frame was predicted,the predicting results was used to correct the tracking results generated by Camshifi.To overcome the interference factors,different algorithms were integrated.To reduce the interference of shadow,the chrominance values of the imager was replaced by the luminance value of the background image; to overcome the impact of light,the discrepancy of H component between HSV image and each frame was used to update the mean weight of H component.The results showed that the improved algorithm can track the bamboo efficiently,quickly and accurately.

关 键 词：计算机视觉 改进Camshift 目标检测 目标跟踪 

分 类 号：S126[农业科学—农业基础科学]