电脑知识与技术
電腦知識與技術
전뇌지식여기술
COMPUTER KNOWLEDGE AND TECHNOLOGY
2009年
21期
5751-5754
,共4页
遗传算法%小波神经网络%图像特征提取%K-L变换
遺傳算法%小波神經網絡%圖像特徵提取%K-L變換
유전산법%소파신경망락%도상특정제취%K-L변환
genetic algorithm%wavelet neural network%image feature extraction%K-L transform
该文利用单层有限神经元的遗传算法和小渡神经网络相结合,以小波网络的速度得到提高;同时注意到K-L的降维、压缩等先进性,遗传算法的鲁棒性,进而将遗传算法、小波神经网络、图像特征提取三者有机结合,促使实时系统能够更快速提取图像特征,同时对图像进行小波压缩和K-L压缩,使压缩率更高.
該文利用單層有限神經元的遺傳算法和小渡神經網絡相結閤,以小波網絡的速度得到提高;同時註意到K-L的降維、壓縮等先進性,遺傳算法的魯棒性,進而將遺傳算法、小波神經網絡、圖像特徵提取三者有機結閤,促使實時繫統能夠更快速提取圖像特徵,同時對圖像進行小波壓縮和K-L壓縮,使壓縮率更高.
해문이용단층유한신경원적유전산법화소도신경망락상결합,이소파망락적속도득도제고;동시주의도K-L적강유、압축등선진성,유전산법적로봉성,진이장유전산법、소파신경망락、도상특정제취삼자유궤결합,촉사실시계통능구경쾌속제취도상특정,동시대도상진행소파압축화K-L압축,사압축솔경고.
In this paper, we will combine with single-layer neurons of the genetic algorithm and wavelet neural network, such that the speed of wavelet network is improved. We note that the K-L dimensionality reduction, compression and other advanced, robust genetic algorithms. We will combination with genetic algorithm, wavelet neural networks and image feature extraction, make real-time system to be able to extract image features more quickly, at the same time wavelet image compression and KL compression for compression ratio is higher.
