Cloud based applications are rapidly developed in recent years and have reshaped models and algorithms. They can offer a user experience like a program installed entirely on a local machine, but with reduced resource needs, more convenient updating, and the ability to access functionality across different devices. A cloud based real-time & wheat spike object and disease detection scheme is being developed to achieve plant object detection and segmentation through a smartphone application. The detection platform are divided into two parts -  detection and application. For wheat spike and disease detection, artificial intelligence (AI) models were trained on a wheat dataset collected from the wheat trial field from Saint Paul, MN. Preprocessing of images were conducted including augmentation and resizing. A Yolov5Tiny model with 87.1 mean Average Precision (mAP) was chosen to detect wheat spikes. For application, a user interface was designed, and functions were developed for Android platforms. The android application uploads images obtained from the smartphone camera or stored in the smartphone to the cloud server for detection modeling. Amazon Web Service (AWS) was chosen as the cloud computing platform, and Microsoft Azure was chosen as its backup. By integrating the Yolov5Tiny model and the Android application, we expect to develop a smartphone application for near real-time plant object detection on users’ Android smartphones.