Obstacle Detection System Based On Civil Aircraft And Implementation

Fatal flight accidents most often occur during the taking off and landing phases, where more understanding of the environment is needed for the pilots. However, lack of low cost and accurate obstacle detection system means that General Aviation (GA) aircraft and civil helicopters have relatively high accident rates when compared with commercial flights which can afford the more expensive sensor equipment. One of the methods to improve this situation is using common software and data fusion techniques combining the data from several low cost sensors in order to get the equivalent performance of using a better specified and more expensive sensor.This paper focuses on the performance of fusing data from infrared and MMW images and was achieved via four discrete steps: image generation, pre-processing, image processing and obstacle detection. Gaussian filter and median filter were used for removing noise in the pre-processing section. Pixel adding and pyramid algorithms were applied to achieve data fusion plus additional processing, such as edge detection and corner detection were used to increase the effectiveness of the method. Morphological processing and a flood filling algorithm were implemented in the final part for obstacle detection.Comparing the results from data fusion technique with that the output from the single data source, it is clear that the former gives much better performance than the latter. This supports the assertion that by fusing the data from different sensor equipments using advance DSIP methods, the output image will provide better performance than individual sensors, and will equal or even surpass the performance of more expensive devices.