Assuming what you're doing is based on this conversion:
I'll describe what you can do in pseudo-code:
First extract x, y from public key.
// get x and y from Public key "point" EC_POINT_get_affine_coordinates_GFp(group, pub_key, x, y, ctx); // convert BIGNUMs x, y into binary form BN_bn2bin(x,x_char); BN_bn2bin(y,y_char); Next you need to do message digest several times, including 3 sha256 and 1 ripemd160. In the following pseudo-code I'll show you how to do ripemd160. To do sha256 with EVP_MD, just replace EVP_ripemd160() with EVP_sha256(), and update (input to EVP_MD) your input message with single or several EVP_DigestUpdate().
EVP_MD_CTX ctx; EVP_MD_CTX_init(&md_ctx); EVP_DigestInit(&md_ctx, EVP_ripemd160()); // hdr = 0x04 EVP_DigestUpdate(&md_ctx,hdr,1); EVP_DigestUpdate(&md_ctx,x_char,32); EVP_DigestUpdate(&md_ctx,y_char,32); // put message degest into dgst and set length to dgstlen EVP_DigestFinal(&md_ctx,dgst,&dgstlen); EVP_MD_CTX_cleanup(&md_ctx); Or the easier way, call sha256() and ripemd160() directly. But you need to prepare your input message before calling hash functions sha256() or ripemd160().
The 25-byte binary address is the result of ripemd160, together with the first 4 bytes of 32-byte checksum. You need to find a way to convert it from Base 256 to Base 58. I don't think OpenSSL support that.
