Numerous engineering applications can be viewed as optimization challenges, often requiring computationally intensive objective functions. Evolutionary Algorithms (EAs) have recently gained traction for tackling complex problems across various fields, offering a more robust approach to finding global optima than traditional algorithms. Nevertheless, achieving satisfactory outcomes when using EAs for practical applications typically requires a large number of function evaluations. Several alternative models have been utilized to replace the original objective function, allowing certain EAs to be adapted in order to decrease the required number of function evaluations and address this issue. Although these methods improve efficiency, the original EA's search abilities and the accuracy of the solutions are notably affected by the use of alternative models. To tackle this challenge in image processing, a new evolutionary technique, known as Adaptive Population with Reduced Evaluations (APRE), has been developed. APRE achieves a reduction in function evaluations through two mechanisms: dynamically adapting the population and incorporating a fitness calculation strategy that determines when to calculate or merely estimate new individuals. This enables a significant reduction in function evaluations while maintaining the effective search capabilities of evolutionary approaches. A comparison is made in this chapter between APRE's performance and other methods that were previously created to minimize function evaluations, focusing on its effectiveness as a global optimization algorithm. Several standard benchmark functions, widely used in the EA field, are utilized for the comparison. The best balance between the number of function evaluations and solution accuracy, in comparison to its counterparts, is confirmed by the simulation results of the presented APRE method.