摘要
受到图像破损区域的影响,对风格元素的识别效率和识别精度较低,由此,提出了基于阈值的图像风格元素智能识别方法。通过小波去噪原始图像,获得较优的风格元素表象;利用灰度直方图对图像阈值分割,用统一的灰度来划分成便于计算的二值图像;利用闭运算法模糊消除区域破损地方,获得完整清晰图像;利用灰度图像智能识别技术根据图像形成的谷形区域实现图像风格元素识别。仿真实验结果表明,所提方法在识别精度、误差率和时间效率方面均较优,能够为实际应用提供理论指导。
Due to the influence of the image damage area, the recognition efficiency and recognition accuracy are low. Therefore, a threshold-based intelligent recognition method of image style elements is proposed.The superior style element image is obtained by canceling the original image;use the gray scale histogram to divide the unified gray scale into a binary image;obtain the complete clear image by eliminating the closed transport algorithm.Simulation experimental results show that the proposed method is excellent in identification accuracy, error rate and time efficiency, and can provide theoretical guidance for practical applications.
作者
蒋平
JIANG Ping(Open Education College,Anhui Open University,Hefei Anhui 230022,China)
出处
《德州学院学报》
2022年第2期51-55,共5页
Journal of Dezhou University
关键词
智能视觉
风格元素
灰度直方图
二值图像
归一化处理
intelligent vision
style element
gray histogram
binary image
normalization processing
