Impact of Camera Vibration Frequencies on Image Noise for Unmanned Aerial System Applications

In the transportation systems engineering field, there is a fast growing interest to use small unmanned aerial systems (UAS) for structural inspections. Of particular importance is the development of image processing algorithms to process UAS collected data for transportation infrastructure (e.g., bridges) defect detection/classification purposes. Results from preliminary experiments using image data collected with a UAS have shown the presence of significant image noise, which negatively impacts the efficiency of algorithms to detect/classify structural defects. This research study presents an ongoing effort to understand the relationship between UAS camera vibrations to image noise. Ultimately, the objective is to identify maximum camera vibration frequencies that will result in acceptable image noise levels for effective detection and classification of defects.