Within food production in agriculture, different sensor networks are used to manage and monitor different processes, ranging from the presence of certain plant nutrients in soil to the presence of pests to determine plant health and to increase yields. Some limitations exist on how it is possible to use these sensor devices to monitor plant environments, for example it is not possible to attach or use one sensor for each plant. These monitoring tools and methods need to be scalable and affordable so they can be expedient in future farms. The potential exists to use already gathered data to combine or fuse different data and make sensors that are used in food production more effective and reduce costs in the process and help farmers to make more informed decisions. This literature review study investigates machine learning and computer vision based research that can be used within monitoring in IoT sensor network architectures with an aim to reduce the physical sensor devices needed in data gathering processes and to reduce the overall cost of IoT network implementation in agricultural farms which are growing food. Different machine learning and deep learning architectures in this context already have been implemented successfully to achieve a high accuracy in data forecast and image processing tasks.