Air Digital Camera Image Processing

Aerial reconnaissance is a new integrated detecting technique, which can not be separated from the group of optics, infranics, microwave technique, radar technique, laser technique, computer and image processing technique. The Unmanned Air Vehicle (UAV) aircraft reconnaissance system is an important branch of aerial reconnaissance. It gets information by aerial digital camera, loaded in UAV. In the camera, digital images mostly got from by image sensor, are preprocessesed, compressesed and transmitted them to the data system.When UAV flies several kilometers high with certain velocity, the camera in it takes pictures of the ground uprightness and processes them with high speed. Under this kind of circumstance, the images always are complicated with low brightness contrast. At the same time, because of the air flows, the clouds, the vehicle shakes, the image sensor's faults and the hardware's noises, the images are companied with noises. The images should be preprocessed, such as enhancement and noise elimination.As the sensor's pixel number is increasing, aerial digital camera has higher and higher resolution, and image data is dramatically increasing that brings storing and transmitting problems. Large capacity memory solved the problem of storing great deal of digital images, but the speed of storing and transmitting can not meet the requirement of real-time storing and transmitting the remote sensing image more that has more than 10 million pixels. Real-time storing and transmitting is important to military affairs. In order to reduce image data, images should be compressed before they are stored and transmitted, which is a substantial research work in the field of aerial digital camera.According to the design requirement of the super high resolution CCD aerial digital camera, which will be used in UAV reconnaissance system, this paper studies image processing technique. First, the noises of CCD are studied and eliminated with proper algorithm. Second, under the requirement of real-time storing and transmitting, this paper studies compressing techniques, especially JPEG, improve and programme JPEG algorithm to compress the remote sensing images. Third, this paper studies image file formats, and carries out a kind of image file format, which is including flight latitude, velocity, altitude and longitude.