Aim: The aim of this paper is to create an intelligent decision support system for spraying and fertilization operations in agricultural lands with Agricultural Unmanned Aerial Vehicles. In this system, different pesticides will be sprayed in sparse or undeveloped areas, different pesticides will be sprayed in areas with normal development and different pesticides will be sprayed in areas with intensive development and plant health will be protected. Thus, inadequate spraying or burning of the plant due to over spraying will be prevented. Method: In this paper, a decision support system embedded in the Jetson Nano Devkit Artificial Intelligence Computer is developed for spraying and fertilization operations performed by Agricultural Unmanned Aerial Vehicles (UAVs). Three different image processing algorithms ("RGB Filter Method", " Pixelwise Classification Method" and "HSV Range Filter Method") were tested on 4000 agricultural field images in order to accurately analyze the snapshots obtained from UAVs. Solution time and solution reliability (green capture rate) were considered as two different objectives in this system where the images should be evaluated instantaneously and the amount of pesticide spraying should be changed. According to the image processing results, the working principle of the spraying system, which was created according to 3 different levels (low-medium-high level) and "Pulse Width Modulation" and "Pulse Width Modulation" techniques, was simulated through the "Proteus" simulation programme and transferred to the real system. Results: As a result of the analysis, it was determined that the "RGB Filter Method" is the ideal algorithm that simultaneously meets the objectives of solution time and solution reliability. In the spraying system, the amount of pesticide to be sprayed per second was determined for 3 stages. Finally, it was decided to take an image every 6 seconds from the camera on the APC and to spray accordingly. Conclusion: Thanks to the proposed decision support system, agricultural areas can be sprayed at dynamic levels according to the level of plant needs. In today's world of global warming, soil erosion and water scarcity, the study is expected to make a positive contribution to smart agricultural practices.