發(fā)布時間：2019-06-03 14:25

【摘要】：光學顯微數(shù)碼成像系統(tǒng)中,光學退化是影響圖像質(zhì)量的重要因素。結(jié)合物鏡測試,提出了一種結(jié)合分視場點擴散函數(shù)估計的梯度約束顯微圖像復(fù)原方法,并應(yīng)用于顯微鏡以形成質(zhì)量提升系統(tǒng)。通過物鏡調(diào)制傳遞函數(shù)測試,分視場計算獲取點擴散函數(shù);引入零階范數(shù)梯度約束作為正則化項,設(shè)計基于該約束的快速復(fù)原方法;并采用漸變加權(quán)拼接方法實現(xiàn)分視場復(fù)原結(jié)果的無縫拼接。對不同物鏡下的顯微圖像進行了復(fù)原測試,并采用評價方法進行評估。結(jié)果表明,本文方法給出的顯微圖像復(fù)原效果好、運行速度快,對尺寸為2048pixel×1536pixel的圖像只需1.62s,適用于顯微成像系統(tǒng)。
[Abstract]:In the optical microscopic digital imaging system, optical degradation is an important factor that affects the quality of the image. In this paper, a gradient-constrained micro-image restoration method combined with the point-of-view point spread function estimation is proposed in combination with the objective test, and the method is applied to the microscope to form a mass lifting system. According to the objective modulation transfer function test, the point spread function is obtained by the field-of-view calculation; a zero-order norm gradient constraint is introduced as a regularized term, and a fast recovery method based on the constraint is designed; and a gradual weighted splicing method is adopted to realize the seamless splicing of the partial-view restoration result. The microscopic images under different objective lenses were tested and evaluated using the evaluation method. The results show that the image restoration effect is good, the running speed is fast, the image with the size of 2048 pixels and 1536 pixels is only 1.62 s, which is suitable for the micro-imaging system.
【作者單位】： 杭州電子科技大學電子信息學院;寧波永新光學股份有限公司;
【基金】：國家自然科學基金(61405052) 浙江省科技計劃項目(2017C01033)
【分類號】：TH74;TP391.41

【相似文獻】

相關(guān)期刊論文 前1條

1 王大勇,伏西洋,郭紅鋒,趙博,鄭怡嘉,陶世荃;光學稀疏孔徑系統(tǒng)的成像及其圖像復(fù)原[J];光子學報;2005年10期

，

本文編號：2492000