Noise quantification is crucial pre-processing step in the field of computer vision and image processing aimed at image enhancement for better analysis and interpretation. Due to the sensor limitations, transmission and acquisition noise entered in the digital images. Many Conventional and deep learning methods are available to remove the noise from the digital images. There is lack of methods to quantify the noise; in this paper, we implemented new method to quantify the impulse noise in natural images. With the help of pixel similarity and Euclidian distance, calculate the threshold value to estimate the noise level in the image. Compare to existing method new method outperform well in estimating the noise level.