Generally selecting of an IR detector for practical and commercial application, is difficult. We have to take into consideration several factors such as availability, cost, environmental stability and processing simplicity. These factors are often as important as radiometric performance and may well determine the final selection of a particular detector. Also, selection of the design parameters of a given detector is different according to the application. This work presents the ability to simulate the response of micro-devices using optimization analysis to give researchers insight of the behavior of devices before prototypes are developed as well as allowing the designer to optimize the desired response. We consider the main types of IR thermal detectors, their basic operational principles, and their performance parameters aiming to judge on their suitability for high speed thermal imaging applications. Mathematical models for the resistive bolometers, p n junction bolometers, pyroelectrics, are presented. Based on these mathematical models, the limiting factors of these detectors are identified, and their effect on the performance measures ( e.g., sensitivity and detectivity) is calculated. Unconstrained optimization techniques are then used to optimize the design parameters of an infrared thermal detector for the best detectivity. Next, constrained optimization techniques are used to evaluate the detectivity subject to a lower limit constraint on the frame rate. Our results indicate that the p-n junction diode bolometers offer the best detectivity at low frame rates (30-60 Hz) at room temperature, with a detectivity of 7x10⁸(cmHzW). For high frame rate applications (100-200 Hz), the resistive bolometers offer the best detectivity at room temperature, with a detectivity of 6x10⁸ (cmHzW).