The UAV is equipped with a hyperspectral imager, which combines optical technology, environmental engineering, Internet of Things and artificial intelligence to create an air quality detection system. In addition to using hyperspectral technology to increase the amount of image information, we also integrate deep learning The technology successfully overcomes the shortcomings of the previous air quality detection can only be a small range, higher cost, further increasing the convenience and accuracy of the system.

3D-CNN, the input feature of this model is different from the first two, which belongs to the spectrum space cube. In addition to the frequency spectrum, the feature also includes the spatial morphological features of National Chung Cheng University.

UAV air quality applications, the scope of government and commercial applications that can be used, the government can use this technology to do air quality testing for manufacturers in industrial areas; the commercial applications of its people are also very extensive and can be combined with smart agriculture , Use air quality testing and the established data set to predict the agricultural harvest results and growth environment for farmers to obtain the best harvest with the least cost.