The interconnection between remote sensing satellites and relay satellites can be achieved through inter satellite links （ISL）， which enables timely transmission of space based remote sensing data to the ground and shorten the response time of remote sensing. Thus， the joint scheduling problem of satellite remote sensing and relay is studied. On the basis of describing and analyzing the problem， taking maximizing the total task priority as the objective function， this paper constructs a remote sensing and relay joint scheduling model with constraints of visible time window， service time window and uniqueness. To solve this problem， a joint scheduling optimization algorithm based on the adaptive large-scale neighborhood search framework， the adaptive large-scale neighborhood search based joint scheduling algorithm Ⅱ （ALNS-JS-Ⅱ） is designed. This algorithm uses the task allocation operator to decompose the multi-satellite scheduling into multiple single-satellite parallel scheduling sub-problems， and uses an adaptive method to select all operators. The whole algorithm realizes the effective interaction between remote sensing scheduling and relay scheduling. In order to verify the effectiveness of the ALNS-JS-Ⅱ algorithm， it is compared with the adaptive large neighborhood search algorithm with high coupling of remote sensing and relay， the adaptive large neighborhood search based highly coupled joint scheduling algorithm Ⅰ （ALNS-JS-Ⅰ）， the two-stage independent scheduling algorithm of remote sensing and relay based on the adaptive large neighborhood search， the two stage ALNS （Ts-ALNS）， in various task scenarios. The experimental results show that the ALNS-JS-Ⅱ algorithm is 4.58% and 1.48% better than ALNS-JS-Ⅰ and Ts-ALNS， respectively， in terms of algorithm solution gain， and 20%—30% faster in terms of solving efficiency. In conclusion， the ALNS-JS-Ⅱ algorithm has a better solving ability for the joint scheduling problem of remote sensing and relay resources.