System and method for space object detection in daytime sky images

1. A system for detecting space objects, the system comprising:

a computer system comprising one or more processors configured to execute one or more computer program instructions that, when executed by the one or more processors, effectuate operations comprising:

generating a spatial background-difference image based on a stacked image generated by stacking a plurality of images of daytime sky obtained from a camera system and one or more offsets and gains, wherein the spatial background-difference image is generated by

(1) generating a clean image by adjusting the stacked image to account for the one or more offsets and gains, and

(2) obtaining a spatial background image using a background estimation kernel, wherein the spatial background-difference image is generated by subtracting the spatial background image from the clean image;

obtaining a matched-filter image based on the spatial background-difference image and a point-source function kernel, wherein the matched-filter image comprises an array of pixels;

generating a mask including a plurality of bits corresponding to the array of pixels, each bit of the plurality of bits comprising a first value or a second value based on whether a signal-to-noise ratio (SNR) of the corresponding pixel of the array of pixels satisfies a threshold condition;

detecting at least one space object in orbit based on clusters of bits in the mask having the first value or the second value; and

generating, based on the mask and the spatial background-difference image, output data indicating whether an observation of the at least one space object in orbit has been detected.

2. The system of claim 1 , further comprising:

a camera system comprising:

one or more shortwave infrared (SWIR) sensors; and

one or more thermoelectric coolers (TECs).

3. The system of claim 1 , wherein obtaining the matched-filter image comprises:

performing a convolution of the spatial background-difference image with the point-source function kernel.

4. The system of claim 1 , wherein the operations further comprise:

determining one or more candidate space objects within the matched-filter image based on the mask, wherein the at least one space object in orbit is detected from the one or more candidate space objects.

5. The system of claim 1 , wherein generating the output data comprises:

generating an output image comprising the spatial background-difference image and the observation of the at least one space object in orbit.

6. One or more non-transitory computer readable media comprising computer program instructions that, when executed by one or more processors, effectuate operations comprising:

obtaining a stacked image by stacking shortwave infrared (SWIR) images of daytime sky, wherein the SWIR images are obtained from a camera system comprising one or more SWIR sensors and one or more offsets and gains;

generating a clean image by adjusting the stacked image to account for the one or more offsets and gains;

obtaining a spatial background image using a background estimation kernel, wherein a spatial background-difference image is generated by subtracting the spatial background image from the clean image;

obtaining a matched-filter image based on the spatial background-difference image and a point-source function kernel;

generating a mask based on the stacked image, wherein the mask comprises a plurality of bits, each bit of the plurality of bits comprising a first value or a second value based on whether a signal-to-noise ratio (SNR) associated with the bit satisfies a threshold condition; and

generating output data based on the mask and the spatial background-difference image, wherein the output data indicates whether a space object in orbit has been detected.

7. The one or more media of claim 6 , wherein the spatial background-difference image is generated further based on the stacked image.

8. The one or more media of claim 7 , wherein the output data is generated further based on the spatial background-difference image.

9. The one or more media of claim 6 , wherein the operations further comprise:

performing a convolution based on the spatial background-difference image and a point-source function kernel to obtain a matched-filter image, wherein:

the mask is generated further based on the matched-filter image,

candidate space objects are detected based on the matched-filter image, and

the output data indicates whether the space object in orbit has been detected based on the candidate space objects.

10. The one or more media of claim 9 , wherein generating the output data comprises:

retrieving, based on a location of the camera system when the SWIR images were captured, space object location information indicating a plurality of space objects in orbit; and

determining that at least one of the candidate space objects is a space object in orbit based on at least one confidence measure indicating that the at least one of the candidate space objects is not a false positive, wherein the output data comprises an indication of the space object in orbit.

11. The one or more media of claim 7 , wherein the stacked image comprises an array of pixels, and each pixel of the array of pixels has a stacked numerical value, the operations further comprising:

generating a clean image based on the stacked image, one or more offsets, and one or more gains.

12. The one or more media of claim 11 , wherein the operations further comprise:

obtaining a bias value associated with each SWIR sensor of the one or more SWIR sensors, wherein the one or more offsets comprise the bias value, and wherein the stacked numerical value of each pixel of the array of pixels of the stacked image is adjusted based on the bias value associated with the pixel.

13. The one or more media of claim 11 , further comprising:

obtaining a correlated noise value associated with each pixel of the array of pixels, wherein the stacked numerical value of each pixel of the array of pixels of the stacked image is adjusted based on the correlated noise value associated with the pixel.

14. The one or more media of claim 11 , further comprising:

determining, based on a precomputed function representing pixel response non-uniformity (PRNU) of each SWIR sensor of the one or more SWIR sensors, a gain value of each pixel of the array of pixels of the stacked image, wherein the one or more gains comprise the gain value, and wherein the stacked numerical value of each pixel of the array of pixels of the stacked image is adjusted based on the gain value associated with the pixel.

15. The one or more media of claim 6 , wherein the camera system further comprises one or more thermoelectric coolers (TECs).

16. The one or more media of claim 6 , wherein generating the output data comprises: generating an output image comprising the spatial background-difference image and an observation of the space object in orbit.

17. A method implemented by one or more processors configured to execute one or more computer program instructions that, when executed by the one or more processors, perform the method, the method comprising:

obtaining a stacked image by stacking shortwave infrared (SWIR) images of daytime sky, wherein the SWIR images are obtained from a camera system comprising one or more SWIR sensors and one or more offsets and gains;

generating a clean image by adjusting the stacked image to account for the one or more offsets and gains;

obtaining a spatial background image using a background estimation kernel, wherein the spatial background-difference image is generated by subtracting the spatial background image from the clean image;

obtaining a matched-filter image based on the spatial background-difference image and a point-source function kernel;

generating a mask based on the stacked image, wherein the mask comprises a plurality of bits, each bit of the plurality of bits comprising a first value or a second value based on whether a signal-to-noise ratio (SNR) associated with the bit satisfies a threshold condition; and

generating output data based on the mask and the spatial background-difference image, wherein the output data indicates whether a space object in orbit has been detected.