Image processing boards often use multi-core-based CPUs to process massive data for two-dimensional signal processing, considering the amount of computation required. Image processing boards that use hardware structures utilizing CPUs and FPGAs to operate image processing algorithms in real time are generally utilized. CPUs have the advantage of being able to perform real-time image processing based on multi-cores, but they have the disadvantage of not being able to perform tasks while the CPU is booting. It is not appropriate to use it in a system that requires rapid operation after power-on and target detection through image processing. We propose a hardware structure utilizing DSPs and FPGAs as a method to shorten the boot time so that it can operate quickly after power-on. A tracking algorithm that boots quickly and processes images is implemented through four DSPs and FPGAs. This structure can provide more flexibility in software development with reconfigurability and expandability for additional image processing.