This paper is a meek venture to introduce the compressed sensing magnetic resonance image reconstruction. I have devised this algorithm by minimizing total variation (TV) and L1 norm regularization with total variation denoising. MR image reconstruction has gained a tremendous far-reaching impact in the present technologically advanced world. The original problem will be bifurcated into two entities; L1norm and Total Variation [1,2]. Eventually it can be effectively solved. It helps the expedite reconstruction of the MR image through an iterative framework. An additional application of a denoising technique to this method was found to be very efficient and reliable. A comparative view of the computational complexity and reconstruction accuracy of the present method with the earlier approaches will open our eyes to the effectiveness of it based on the numerical results