The multi-constrained QoS-based routing problem of wireless multimedia sensor networks is an NP hard problem. Genetic Algorithms (GAs) have been used to handle these NP hard problems in wireless networks. Because the crossover probability is a key factor of GAs’ action and performance, and affects the convergence of GAs, and the selection of crossover probability is very difficult, so we propose a novel method – crossover game instead of probability crossover. The crossover game in routing problems is based of each node has restricted energy, and each node trend to get maximal whole network benefit but pay out minimum cost. The players of crossover game are individual routes. The individual would perform crossover operator if it is Nash equilibrium in crossover game. The Simulation results demonstrate that this method is effective and efficient.