Photography on smartphones is getting more advanced and quality by using higher megapixels, resulting in higher quality images. The use of smartphones has now become a lifestyle for the wider community. Photographic results are not only easy to remember, but sometimes have a very crucial role and are always interesting to review and observe because they have a very high documentation value. A photo is able to record something that may not be repeated, it can be in the form of events or human activities, natural atmosphere, pictures of personal life or natural events and cultural arts in society. The problem in this research is how to effectively reduce noise with arithmetic mean filter method on low specification smartphone cameras and how to apply arithmetic mean filter algorithm on low specification smartphone cameras. The type of research used is a type of waterfall. The results showed that the application can reduce salt & pepper noise well. The arithmetic mean filter method can eliminate noise better than the harmonic mean filter method when taking pictures on a smartphone camera by testing the MSE (Mean Square Error) and PSNR (Peak Signal to Noise Ratio) values with the MSE arithmetic mean filter value of 0.48 while the MSE harmonic mean filter is 0.56 and the PSNR arithmetic mean filter is 97.976 dB, while the PSNR harmonic mean filter is 96.637 dB.