METHOD AND DEVICE FOR REMOVING IMAGE NOISE BASED ON THE MEDIAN FILTER

A method and an apparatus for image denoising based on medium filter is disclosed according to the present invention. Such method and apparatus may be used to remove linear noise and random noise for the sensor, as well as the bad points in the sensor image, and to prevent over-denoising. The technical solutions includes selecting three groups of pixels from neighboring pixels of a current pixel, calculating medians Φ1, Φ2 and Φ3 for the pixels in each group; calculating a reference luminance Y: Y=Φ2+[Φ1+Φ3−2*Φ2]*h*[1−k*var], where var=Φ3−Φ1; determining whether the current pixel (cur_pixel) is a bad pixel; letting a temporary variable (tmp_data) be Φ1 if the current pixel is a bad pixel; letting a temporary variable (tmp_data) be the current pixel value (cur_pixel) if the current pixel is not a bad pixel; determining whether the current pixel is located on an edge. If the current pixel is determined to be located on the edge, the reference luminance Y is limited to a range of [tmp_data−edge_limit, tmp_data+edge_limit], the current pixel value is updated with the limited reference luminance Y and the denoising process ends. If the current pixel is determined not to be located on the edge, the value of |Y−tmp_data| and the value of avoid_over_noise are compared. If |Y−tmp_data|<avoid_over_noise, the current pixel value is updated with Y; otherwise, the current pixel value is updated with tmp_data. The above steps for each pixel in the image is repeated. The present invention is applicable to image denoising field.