本文關(guān)鍵詞： 激光束 高光譜圖像 數(shù)據(jù)存儲　出處：《激光雜志》2017年04期 　論文類型：期刊論文

【摘要】：針對激光束干擾情況下,高光譜圖像存儲過程中存在的丟幀問題,提出基于稀疏表示與隊(duì)列式緩存相結(jié)構(gòu)相結(jié)合的高光譜圖像數(shù)據(jù)存儲方法研究。首先利用Gabor濾波器組,濾除激光束對高光譜圖像的干擾,提高光譜圖像的可識別度;基于稀疏表示原理構(gòu)造訓(xùn)練樣本集合,利用獲取的基函數(shù)字典對高光譜圖像數(shù)據(jù)進(jìn)行稀疏編碼,完成對圖像的壓縮處理;最后基于隊(duì)列式緩存結(jié)構(gòu)消除存儲速度波動帶來的不良影響,實(shí)現(xiàn)高光譜圖像數(shù)據(jù)的實(shí)時存儲。實(shí)驗(yàn)證明了提出高光譜圖像數(shù)據(jù)存儲方法,能夠有效提高存儲速度,并解決了圖像存儲過程中的數(shù)據(jù)丟失問題。
[Abstract]:Aiming at the problem of losing frames in hyperspectral image storage under laser beam interference, a method of hyperspectral image data storage based on sparse representation and queue buffer structure is proposed. Firstly, the Gabor filter bank is used to store hyperspectral image data. Filter out the interference of laser beam to hyperspectral image, improve the recognition degree of spectral image, construct the training sample set based on sparse representation principle, and use the obtained basis function dictionary to encode the hyperspectral image data sparsely. Finally, based on the queue buffer structure, the bad effect of storage speed fluctuation is eliminated, and the real-time storage of hyperspectral image data is realized. The experimental results show that the hyperspectral image data storage method is proposed. It can effectively improve the storage speed and solve the problem of data loss in image storage process.
【作者單位】： 包頭師范學(xué)院數(shù)學(xué)科學(xué)學(xué)院;
【基金】：內(nèi)蒙古教育廳項(xiàng)目(NJSG201282)
【分類號】：TP751

【相似文獻(xiàn)】

相關(guān)期刊論文 前1條

1 王露露;王利;胡增康;;基于閃存卡的視頻圖像數(shù)據(jù)存儲方法[J];探測與控制學(xué)報(bào);2012年03期

，

本文編號：1503049