An extraction method of color similarity feature in the rotating moving image is proposed based on the Photoshop technology.Firstly,the pixel color value of rotating moving image is expressed as the similar histogram form by using the Photoshop technology,and the image area is divided.Then,the association degree of color overall variation rule between any two points in the area is calculated to obtain the feature vector,and the vector weight of two kinds of division area is obtained through the iteration processing according to the similarity maximization rule.Moreover,the feature extraction of color similarity in the rotating moving image is converted into the weight optimization problem.Finally,the extraction of color similarity feature is completed according to the obtained optimal solution.The experimental results show that the method can improve the extraction precision.It reduces the influence of illumination intensity variation on the color and has strong robustness.%在对旋转移动图像的颜色相似度特征提取过程中,图像颜色本身具有不稳定性,使得初始特征向量提取不准确.传统方法需要先获取图像上各个网格的颜色直方图,得到图像的初始特征向量,但容易受到外界因素影响,局限性大,存在提取误差大的问题.提出基于Photoshop技术的旋转移动图像颜色相似度特征提取方法.应用Photoshop技术将旋转移动图像的像素颜色值表示为类似直方图的形式,对图像区域进行划分,计算区域内任意两点之间颜色整体变化规律的关联度,得到特征向量,依据近似性最大化准则,通过迭代处理获得两类划分区域的向量权重,将旋转移动图像颜色相似度特征提取问题转换为权重的优化问题,依据获取的最优解完成对图像颜色相似度特征的提取.实验结果表明,所提方法能够有效提高旋转移动图像颜色相似度特征提取精度,减少光照强度变化对颜色的影响,具有较高的鲁棒性.
展开▼