The proposed method introduces impulse noise removal on the image sequences using Dual Recursive Hybrid Filter. The hybrid filter is combined from the Rank Ordered Mean (ROM) filter and Mean-of-Gradient Filter (MGF). The ROM is applied in condition of image sequence at the smallest noise density. Meanwhile, the MGF is applied for the high impulse noise density. An effective method for maintaining an important pixel data from the image sequence has been obtained in the Dual Recursive Hybrid Filter. The qualitative and quantitative parameter of the proposed method provides a good performance filter in many variation densities.
（英）
The proposed method introduces impulse noise removal on the image sequences using Dual Recursive Hybrid Filter. The hybrid filter is combined from the Rank Ordered Mean (ROM) filter and Mean-of-Gradient Filter (MGF). The ROM is applied in condition of image sequence at the smallest noise density. Meanwhile, the MGF is applied for the high impulse noise density. An effective method for maintaining an important pixel data from the image sequence has been obtained in the Dual Recursive Hybrid Filter. The qualitative and quantitative parameter of the proposed method provides a good performance filter in many variation densities.