摘要
针对目前数字图像目标识别方法中存在识别精度和实时性的问题,提出一种结合Gabor小波和神经网络的图像目标识别方法.该方法首先对图像进行预处理,用Canny算子进行边缘提取,然后通过神经网络获取最优的双Gabor小波复合滤波器参数,再采用参数优化过的滤波器组提取目标的特征向量,最后进行目标的分类和识别.实验表明这种方法鲁棒性好、识别率高,具有较广泛的实际应用价值.
At present, many algorithms for object recognition have a narrow applicability and a low effectiveness. To address this issue, this paper presents an image recognition method based on Gabor wavelet and neural network. The images are pre-processed and the edges of objects are extracted with Canny operator first. Then, the combination of Gabor wavelet and neural network is used to obtain the best parameters of dual-compound Gabor wavelet filters. At last, the filter of which parameters has been optimized is applied to extract the eigenvector of the target for object classification and identification. Experiments show that this approach is robust, with high recognition rate and a wide range of practical application.
出处
《扬州大学学报（自然科学版）》
CAS
CSCD
北大核心
2009年第2期49-52,74,共5页
Journal of Yangzhou University：Natural Science Edition
基金
国家自然科学基金资助项目(20299030)
扬州大学自然科学基金资助项目(KK0313090)
