摘要
针对Harris角点检测算法存在对含有噪声的图像角点检测不准确的问题,为降低噪点对角点检测的干扰和提高角点检测的精准度,提出了一种改进算法,通过在Harris算法设置的局部移动窗口中加入一种加权系数自适应的加权均值滤波,其加权系数由像素灰度值决定。通过实验证明,该方法可有效减少伪角点的形成,同时具有良好精确度和检测速度。
In view of the Harris corner detection algorithm has an inaccurate problem in detecting corners of the image whichhas noise, in order to reduce the interference of noise when detecting corners and improve the accuracy of corner detection, animproved algorithm was proposed by adding a weighting coefficient adaptive weighted mean filter to the local motion window set bythe Harris algorithm, the weighting coefficient is determined by the pixel gray scale value. Experiments show that this method caneffectively reduce the formation of pseudo？corner points and has good accuracy and detection speed.
出处
《仪表技术与传感器》
CSCD
北大核心
2017年第12期107-109,104,共4页
Instrument Technique and Sensor
关键词
HARRIS算法
特征点检测
角点检测
加权系数自适应
图像处理
Harris algorithm
feature point detection
corner detection
weighting coefficient adaptive
image processing
