The new generation of smartphones, equipped with various sensors, such as a three-axis accelerometer, has shown potential as an intelligent, low-cost monitoring platform over the past few years. This paper reports the results of an analytical and experimental study on a proposed SDOF model-based noisy deconvolution (SMND) coupled with a deechoing technique to estimate pavement profiles and to modify their geometry using a smartphone inside a vehicle. In the analytical study, the acceleration response of the car was obtained, where the input was a road profile with an arbitrary pattern. Two different methods, classical band-pass filter and wavelet-denoising technique, were used for denoising the acceleration response. In a 2-step deconvolution process coupled with a deechoing technique, the pavement profile was extracted and compared with the original pavement profile, demonstrating good agreement. In the next step, a parametric study was performed to evaluate the effect of vehicle characteristics and speeds. Then, a case study was conducted in Blacksburg, VA, to evaluate the capability of the proposed method in identifying profile types such as potholes and speed bumps. The acceleration-versus-time responses in vertical direction were recorded using smartphone accelerometers located in a moving vehicle. Then, the proposed approach was applied to remove the echo and vehicle dynamics effects to obtain the pavement profiles and to modify their geometry. The results showed that the proposed approach can remove the echo and vehicle dynamics effect from the response to obtain the pavement profile even if the vehicle characteristics and speed are changed.