【摘要】 昆虫翅膀边缘是昆虫翅膀运动变形三维重构的主要依据。通过分析翅膀边缘图像的特点,提出了一种用于提取昆虫高速运动图像亚像素级翅膀边缘的图像处理方法。该方法首先采用高斯滤波预处理图像,然后利用模糊小波增强算法增强图像的弱边缘信号,再通过求灰度图像的梯度,在其梯度变化方向求极值点的方法提取翅膀边缘特征。文中还给出小波原理检测边缘的翅膀边缘检测结果。最后,给出几种算法实验结果的比较和分析,表明该方法成功实现了弱昆虫翅膀的边缘特征提取,并有效抑制了噪声和干扰。更多还原

【Abstract】 An image processing method to extract sub-pixel wing edges for the images of high-speed insects is proposed.With the method,sub-pixel positions of weak wing edges are extracted as follows:the noises are eliminated using gauss filtering at first,then the wing edges are enhanced using fuzzy-wavelet enhancement method,finally the image gray gradient is calculated,and the gradient extremum is detected along the gradient change direction.Also the wing edges detection result is presented by wavelet theory.The experimental result of the proposed approach is compared with those of other methods and analyzed in the end,which demonstrates that the proposed approach extracts wing edges successfully,restrains noise and interference efficiently.更多还原