<p>Tribal art is a very old and ethnic cultural heritage of rural India particularly of Santhal Pargana region, along with its own regional and diversified identity. Tribal artworks are tradsitionally crafted using natural materials that are culturally significant, but highly prone to degradation. These materials are vulnerable to the environmental factors, hence needs proper restoration and preservation. Traditional restoration methods are time consuming and require skilled artisans, which limits the number of artworks that can be restored manually. So, in this paper we have used digital restoration of distorted artworks using varied image processing techniques. Different histogram equalization (HE) techniques namely BBHE (Brightness Preserving Bi-Histogram Equalization), DSIHE (Dualistic Sub-Image Histogram Equalization) and modified BBHE along with various filters have been used here for restoration of distorted tribal artworks. As an extension of the investigation, deep learning based Convolutional Neural Networks (CNN) approach has been used for image restoration purpose. Parameters like AMBE, PSNR, NMSE, CPP and EME have been calculated to find out the quality of the restored images. Results show significant quality improvement of the reconstructed image. The results discussed here may be very much useful for the efficient restoration and preservation of the distorted tribal art works depending on the quality of the sample image.</p>