摘要
根据海上监控采集的海面图像特点,提出了一种基于海天线旋转后准确确定目标区域的方法。该方法是,将海天线图像进行灰度转换,求取二值化阈值,进行图像二值化处理;将灰度图像进行改进的Hough变换得到海天线,旋转灰度图像使海天线呈水平位置;计算海天线在旋转后图像中的高度,根据该数值将旋转后的灰度图像划分为大海、海天交界区和天空区域;对灰度图像再次进行二值化处理;对图像进行反白处理和形态学闭运算,求取连通区域,目标最大的外接矩形即为目标所在区域。实验表明,该方法的检测精度高,同时能避免漏掉待检测目标和产生虚假目标的可能,具有较强的实用性。
Based on the characteristics of sea surface images collected by sea monitoring, a method based on accurate determination of target area after sea antenna rotation is proposed. The method is that the sea antenna image is gray-scale converted to obtain the binarization threshold, and the image binarization is carried out. The Hough transform of the gray image is improved to obtain the sea antenna, and the rotating gray image makes the sea antenna horizontal position;Calculating the height of the sea antenna in the post- rotation image, dividing the rotated gray image into sea, sea-day boundary and sky area according to the value; Binarizing the gray image again;And morphological closed operation, to obtain connectivity area, the largest target external rectangle is the target area. Experiments show that the method has high accuracy and avoids the possibility of missing targets and false targets. It has strong practicability.
出处
《自动化与仪器仪表》
2017年第11期33-35,共3页
Automation & Instrumentation
基金
江苏省重点资助项目研发计划(产业前瞻与共性关键技术)(BE2016009-1)
关键词
海天线旋转
二值化
HOUGH变换
反白处理
形态学闭运算
sea antenna rotation
binarization
hough transform
anti-white processing
morphological closed operation