Using the initial calibration data, the top of atmospheric no-uniformity coefficient for raw data can be drawn as:

D N L = K i , j ⋅ B L ( i , j ) ¯ + C i , j (3)

D N H = K i , j ⋅ B H ( i , j ) ¯ + C i , j (4)

B L ( i , j ) ¯ = 1 m ⋅ n ∑ i = 1 m ∑ j = 1 n B L ( i , j ) (5)

B H ( i , j ) ¯ = 1 m ⋅ n ∑ i = 1 m ∑ j = 1 n B H ( i , j ) (6)

Above processes fit response of every pixel to the statistical mean value, and generate the initial gain and offset coefficients K i , j and C i , j . Figure 4 shows the correction result in lab.

Specially, for Bayer image, there should be two normal approaches for RGB channels calibration separately:

1) interpolate Bayer to virtual RGB, calibrate RGB separately;

2) pick real RGB from Bayer, calibrate them and demosaic to get final result.

We found the latter one is a physical and strict data reconstruction method which generates better result. In one word, Bayer image cannot be interpolated to RGB before radiometric calibration. Figure 5 illustrates the specific operation.

After the launch, non-uniform noise still residues in real image, and it can be easily attributed to atmosphere or some other transfer jitter. Actually, with the help of our former work [7], we stacked the high DN image from mid latitude cloud and low level image covering nightly ocean, and got near 3% radiometric result after a second calibration. Figure 6 outlines the strengthened result from our method.

Excellent results can usually be achieved with the inexpensive, widely-available 24-patch X-Rite color checker. The CCM (color correction matrix) can be applied to images to achieve optimum color reproduction, which is defined as the minimum mean squared color error between a corrected test chart patches and the corresponding reference values, which can be the standard (published) values.

From the radiometric model, the R, G, and B values of each output (corrected) pixel is mathematically a linear combination of the three input color channels, Equation (7) shows the recommended 3 × 3 CCM A and its usage, where the P is processed pixel and O is the original color value:

[ P R 1 P G 1 P B 1 P R 2 P G 2 P B 2 ⋯ P R k P G k P B k ] = [ O R 1 O G 1 O B 1 O R 2 O G 2 O B 2 ⋯ O R k O G k O B k ] [ A 11 A 12 A 13 A 21 A 22 A 23 A 31 A 32 A 33 ] (7)

And also, 4 × 3 CCM usually works fine, please refer to [8] for more details. Figure 7 is the picture of X-Rite color checker from OVS-1A and its correction result.

Consider the launching time is near winter in China, the atmosphere is somewhat cloudy especially near the North China, we introduce the dark channel prior to remove the bad effect and maintain a comfortable visualization. Figure 8 lists the demo case for Zhengzhou airport.

Due to the mechanical vibrations caused by the engine and the effect of turbulent

wind, the video is distorted and unstable. Therefore the video frames have to be stabilized for immediate object detection and real time surveillance. Video stabilization is a process to reduce the unwanted motion of the camera so as to remove the jitter and instability in the video frames [9].

Mechanical stabilization employs mechanical devices like gimbal to reduce the vibration of the structure holding the camera. However, this method is not able to reduce all the vibrations and is costly to implement. The Digital stabilization technique, however employs image processing based methods and can be used to remove the needless vibration that cannot be removed by Mechanical stabilization method. It costs less than the Mechanical technique as it does not involve complex hardware and requires only software to stabilize the video. Therefore digital stabilization is a better method of stabilization since it is inexpensive and easy to implement.

We noticed a parametric strategy in IPOL (www.ipol.im) with open source implementation [10], and used feature-based motion (homography) estimation with local matrix-based motion smoothing, and Figure 9 shows the basic framework.

For further approximately still video, we choose a registration method based on feature point to generate such product. SIFT [11] or ASIFT [12] feature descriptor can be used directly after receiving the stabilized sequences. Figure 10 shows the processed result for Marseille.

Although the compression standard is common in close range community, we got response from end user the compression always damage the expert analysis for small object, especially the high profile H.264 standard (it will incur small blocks near the edge and is of various compression ratios).

Hence, we limited the compression ration no more than 10 for interested region (or quality test video) and no more than 30 in other regions.

After building RPC file using attitude and GPS data, we can construct the L2 product mapped into a cartographic projection and corrected from sensor and terrain distortions. The mean value of geometrical location error is 500m after system correction and calibration (small satellite without high level star sensor).

Mosaic products will be automatically processed in order to provide end user with an ortho-image of larger size, this is vital for multi footprint mode.

We do not pave the traditional knife edge pattern (high cost) but find some natural object such as road or building to test the MTF value, showing in Figure 11.

Table 2 simply outlines the SNR result testified in different condition. Obvious difference can be seen the SNR value is not high while 45 db is common for TDI (Time delay integration) sensor, and explanation is the frame exposure time will not be long to balance DN value and fast moving object detection.

To quantify our video stabilization method, we carefully detect the matched feature points and calculate the residual error of every frame. In Figure12, it can be seen the average of error distance is reduced from largest 5 pixels to no more than 1 pixel in every frame after our processing, which prove its effectiveness (please go to Table 3).

One thing should be kept in mind, this approach will not be used in video sequences with rare feature point, such as small ocean island, desert region.