Based on my basic undestanding, to calculate the PSNR and SSIM, you require of two images-
First Image 1, which is orginal degrded/compressed image and Image 2, whcich is reconstructed from image 1.
Although a higher PSNR generally indicates the reconstruction method is of higher quality, its having some dB value like, more than 20 dB is acceptable (Again it is varried in different cases), I dont think, there is any sepecific values of PSNR, its depends with type of problem.
On the other hand, Structural SIMilarity (SSIM) index is a method for measuring the similarity between two images. The SSIM values ranges between 0 to 1, 1 means perfect match the reconstruct image with original one.
Generally SSIM values 0.97, 0.98, 0.99 for good quallty recontruction techniques.
Request you to read the PSNR and SSIM wikipedia), and if there is clarification needed, please mention the question in comment section as below.
Hope it Helps!
