This paper proposes a new weighted mean filter with a two-phase noise detector for image denoising. Operations are carried out by the detection followed by filtering strategy. For detection, a two-phase noise detector is presented to detect impulse noise (IN). In the first phase, a rank-ordered difference of ROAD (ROD-ROAD) scheme is introduced for detecting noise candidates. Different from most of the existing IN detectors, the proposed detector identifies a pixel by a fuzzy rule that matches the stochastic nature of IN and greatly improves the denoising performance. In the second phase, a local image statistic minimum edge pixels difference (MEPD) is proposed to identify edge pixels from noise candidates. This preserves edges from being wrongly detected as noise; therefore, improves the detection accuracy. For filtering, we design a new weighted mean filter (WMF) that is more suitable for IN to suppress the detected noisy pixels. Finally, an iterative denoising algorithm is presented by combining the proposed two-phase noise detector and the new WMF. The proposed method is accessible and easy to implement. Experimental results show that the proposed method outperforms all the tested state-of-the-art denoising methods with respect to the visual effects and quantitative measure results. © 2015 Elsevier Inc. All rights reserved.