C++ :: JPEGs Recovery From Forensic Image - Optimizing Code

I'm taking a CS course and we've been tasked with creating a program that recovers jpegs from a formatted CF card which uses the FAT file system with a block size of 512 bytes, the jpegs in the card are block aligned which means that the beginning of a jpeg marks the end of the former.

I've wrote the program and it works nicely and recovers the 51 jpegs in the CF card (actually just an image of some 4-5 megabytes of the actual card which can be downloaded here) but I'm looking for ways to optimize/improve my code so I need a second look from experienced programmers.

Here's my code:

Code:
/*
* filename : recover.c
* description : Recovers jpegs from a forensic image
*/

#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <stdbool.h>
#define BLOCK 512 // fat file system block size
int main(void) {
// open CF card image

[Code] .....

My questions are:
1) What are the possible optimizations that can make my code faster/better/more concise?
2) I'm not handling slack space, since trailing zeros at the end of a jpeg won't affect rendering it (we were told it won't contain any garbage values)... how may I approach the problem?
3) Can I enhance my if else constructs? or even replace them with something more elegant?