In the classical deconvolution based on wave-front sensing (DWFS), point spread function (PSF) is considered to be accurately achieved and therefore Wiener filter can be used in retrieval. However, for the inevitable noise with PSF, the object image cannot be restored perfectly. A myopic deconvolution method based on joint estimation of the object and PSF to overcome this limitation is proposed. It takes into account the available a priori information on the object to be restored as well as on PSF. The a priori information is used as constraint of the object and PSF, and therefore a better estimation of the restored object can be obtained. The principles and experimental procedure are presented. The experimental results for the indoor point source show that the myopic DWFS can give a better object restoration than Wiener filter method.