Multi-scale plant phenotyping technologies are capable of collecting big vision-based and spectroscopic datasets, based on which reliable phenotypic analysis can be carried out through a range of computational algorithms based on computer vision and machine learning techniques. Quantitative traits measurement is key to crop genetics, breeding, cultivation and agricultural practices, because they can be used to dynamically evaluate yield, quality and stress resistance in a high-throughput and reproducible manner. As an important staple crop in China, it is essential to establish a systematic approach to monitor wheat growth and quantify yield-related traits during key growth stages. In this work, we firstly reviewed important yield-related traits for bread wheat and then developed a field phenotyping approach to collect a number of common traits using cost-effective and low-altitude unmanned aerial vehicles (UAVs). Based on the visible spectrum images acquired in a field experiment, we utilized professional software (i.e. Pix4Dmapper) to stitch UAV sub-images as well as to reconstruct 3D point cloud to represent the whole experimental field. After this phase, we developed an automated traits analysis pipeline to produce the vegetation map (e.g. Excess-Green index, ExG) and measure important yield-related traits. We have quantified plant height, vegetation index (e.g. ExG) and leaf area index at five key growth stages for 18 wheat genotypes. Our work validates that yield-related traits can be acquired through cost-effective UAVs, which can lower the threshold of conducting field phenotyping and reliable phenotypic analysis. Our work also exhibits a promising approach for research groups and organizations to follow standardize data collection, phenotyping data ontology, as well as the utilization of open-source analytic libraries to develop high-throughput phenotypic analysis techniques in crop phenotyping research.