五邑大学学报（自然科学版）
五邑大學學報（自然科學版）
오읍대학학보（자연과학판）
Journal of Wuyi University(Natural Science Edition)
2015年
3期
51-56
,共6页
改进脉冲耦合神经网络%四方向差值绝对最小%中值滤波
改進脈遲耦閤神經網絡%四方嚮差值絕對最小%中值濾波
개진맥충우합신경망락%사방향차치절대최소%중치려파
improved pulse coupled neural networks (IPCNN)%four-directional minimal absolute difference (FDMAD)%median filtering
为了更好地去除图像噪声,基于图像有效轮廓及边缘连续性,结合改进脉冲耦合神经网络模型(IPCNN),提出了四方向差值绝对最小滤波算法(FDMAD). 仿真实验表明:对比传统基于迭代脉冲耦合神经网络中值滤波方法,本文算法具有更好的去噪效果和更快的计算速度.实验结果验证了算法的快速性和有效性,具有广泛的应用前景.
為瞭更好地去除圖像譟聲,基于圖像有效輪廓及邊緣連續性,結閤改進脈遲耦閤神經網絡模型(IPCNN),提齣瞭四方嚮差值絕對最小濾波算法(FDMAD). 倣真實驗錶明:對比傳統基于迭代脈遲耦閤神經網絡中值濾波方法,本文算法具有更好的去譟效果和更快的計算速度.實驗結果驗證瞭算法的快速性和有效性,具有廣汎的應用前景.
위료경호지거제도상조성,기우도상유효륜곽급변연련속성,결합개진맥충우합신경망락모형(IPCNN),제출료사방향차치절대최소려파산법(FDMAD). 방진실험표명:대비전통기우질대맥충우합신경망락중치려파방법,본문산법구유경호적거조효과화경쾌적계산속도.실험결과험증료산법적쾌속성화유효성,구유엄범적응용전경.
In order to better remove image noise, a Four-Directional Minimal Absolute Difference Filtering algorithm is proposed based on the contours and edge continuity of images. Simulation experiments show that compared with the traditional median filtering method based on iterative pulse coupled neural network, the proposed algorithm has better denoising effect and faster calculation speed. The experiment results verified the fastness and effectiveness and of the Four-Directional Minimal Absolute Difference Filtering algorithm, which has broad application prospects.
