System and Method of a Digital Weapon Sight with Calculated Aimpoint Correction

Abstract

Systems and methods are disclosed to correct aimpoint of digital weapon sight. The method employs the components of a digital weapon sight configured together for a process of sensing of a ballistic event, storing of orientation data and digital imagery data before and after a shot or series of shots, and then performing an image registration algorithm using the imagery data stored. The image registration determines the translational offset between the images, and aimpoint of subsequent shooting is corrected through application of adjusted reticle symbology or adjusted image sensor windowing.

What is claimed is: 
     
         1 . A digital weapon sight comprising a housing, a digital image sensor, a processor, digital memory, and an imagery output mechanism providing a user view of imagery, wherein the processor is configured to:
 store in the digital memory a first imagery data set corresponding to light captured by the digital image sensor,   store in the digital memory a second imagery data set corresponding to light captured by the digital image sensor,   compute a translational offset between the first imagery data set and second imagery data set by performing an image registration algorithm, and   overlay a reticle symbol upon video displayed in the imagery output mechanism according to the translational offset to thereby correct subsequent aiming.   
     
     
         2 . The digital weapon sight of  claim 1 , further comprising a ballistic event sensor, wherein the processor is further configured to detect the occurrence of at least one ballistic event. 
     
     
         3 . The digital weapon sight of  claim 1 , further comprising an orientation sensor, wherein the processor is further configured to: apply a rotational translation to a set of imagery data. 
     
     
         4 . The digital weapon sight of  claim 1 , wherein the processor is configured as part of the image registration to:
 create two frequency-domain data sets by applying a two-dimensional frequency-domain transform on data derived from each of the first and second imagery data,   calculate a complex conjugate of one of the frequency-domain data sets,   multiply a frequency-domain data set with the complex conjugate,   calculate an inverse frequency-domain transform of the resultant product of the multiplication, and   determine the translational offset between the first and second imagery data by finding a maximal value of the magnitude of the inverse frequency domain transform.   
     
     
         5 . The digital weapon sight of  claim 1 , wherein the processor is configured as part of the image registration to:
 create of a plurality of first annular data sets from an annulus of points in the first imagery data,   create a set of second annular data from an annulus of the second imagery data,   calculate the spatial domain correlations between each of the plurality of first annular data sets and the second annular data, and   determine the location of the centroid of the second imagery data within the first imagery data according to comparison of the values of the correlations so calculated.   
     
     
         6 . The digital weapon sight of  claim 1 , wherein the processor is configured as part of the image registration to:
 create of a plurality of first annular data sets from an annulus of points in the first imagery data,   create a set of second annular data from an annulus of the second imagery data,   apply a frequency domain transform on all sets of annular data to create a plurality of frequency-domain data sets,   calculate the complex conjugate of frequency-domain data,   multiply the frequency-domain data sets together,   calculate an inverse frequency domain transform of the resultant product of the multiplication, and   determine the translational offset between first and second imagery data by finding a maximal peak-to-side-lobe ratio for each result of the inverse transform step.   
     
     
         7 . The digital weapon sight of  claim 1 , wherein the processor is configured as part of the image registration to:
 select a two-dimensional patch of data derived from the second imagery data set,   perform a convolutional operation by convolving the patch across data derived from the first imagery data set and, at each position of the patch, calculate a similarity metric, and   find an extremum of all the similarity metrics.   
     
     
         8 . The digital weapon sight of  claim 7 , wherein the similarity metric is calculated by subtracting corresponding data from the two imagery data sets, and the extremum is a minimum value of the similarity metric. 
     
     
         9 . The digital weapon sight in  claim 7 , wherein the similarity metric is calculated using a correlation-based algorithm, wherein corresponding data from the two imagery data sets are multiplied, each datum of the sets having been normalized by a factor derived from a sum of values within its respective imagery data set, wherein that factor is also derived from the number of values in the sum, and wherein the extremum is a maximum value of the similarity metric. 
     
     
         10 . The digital weapon sight of  claim 1 , wherein the processor is further configured to:
 process imagery data by an operator array comprising first or second order partial derivatives or absolute values thereof, or combination of first and second order partial derivatives or absolute values thereof, and   produce by the image registration function a set of two-dimensional coordinates, one for each partial derivative function in the operator array, and   logically analyze these sets of two-dimensional coordinates to produce the translational offset.   
     
     
         11 . The digital weapon sight of  claim 1 , wherein the processor is further configured to:
 determine a difference between data derived from the first imagery data and data derived from the second imagery data, and   logically analyze the values of the difference to determine the location of the point of impact of a projectile.   
     
     
         12 . The digital weapon sight of  claim 11 , wherein the processor is further configured, as part of the determination of the point of impact of a projectile, to:
 detect and remove line segments or edges residual of a difference between data derived from the first imagery data and data derived from the second imagery data.   
     
     
         13 . The digital weapon sight of  claim 2 , wherein the processor is further configured to:
 detect an interrupt from the ballistic event sensor, and   halt imagery data storage based on an interrupt from the ballistic event sensor.   
     
     
         14 . The digital weapon sight of  claim 1 , wherein the processor is further configured to:
 perform a plurality of image registrations,   validate the plurality of image registration algorithms by comparing a plurality of translational offset outputs of image registration algorithms against each other.   
     
     
         15 . An aimpoint correction method, which is applied to a digital weapon sight, comprising:
 storing in the digital memory a first imagery data set corresponding to light captured by the digital image sensor,   storing in the digital memory a second imagery data set corresponding to light captured by the digital image sensor,   computing a translational offset between the first imagery data and second imagery data by performing an image registration algorithm,   and overlaying a reticle symbol upon video displayed in the imagery output mechanism, according to the translational offset, to correct subsequent aiming.   
     
     
         16 . The method according to  claim 15 , further comprising the sensing of a ballistic event. 
     
     
         17 . The method according to  claim 15 , further comprising:
 deriving the rotational difference between two sets of orientation data, and   applying a rotational translation to a set of imagery data.   
     
     
         18 . The method of  claim 15 , wherein the image registration algorithm comprises:
 creating two frequency-domain data sets by applying a two-dimensional frequency-domain transform on data derived from both first and second imagery data,   calculating a complex conjugate of one of the frequency-domain data sets,   multiplying a frequency-domain data set with the complex conjugate,   calculating an inverse frequency-domain transform of the resultant product of the multiplication, and   determining the translational offset between the first and second imagery data by finding a maximal value of the magnitude of the inverse frequency-domain transform.   
     
     
         19 . The method of  claim 15 , wherein the image registration algorithm is a spatial-domain annular method comprising:
 creating of a plurality of first annular data sets from an annulus of points in the first imagery data,   creating a set of second annular data from an annulus of the second imagery data,   calculating the spatial domain correlations between each of the plurality of first annular data sets and the second annular data, and   determining the location of the centroid of the second imagery data within the first imagery data according to comparison of the values of the correlations so calculated.   
     
     
         20 . The method of  claim 15 , wherein the image registration algorithm is a frequency-domain annular method comprising:
 creating of a plurality of first annular data sets from an annulus of points in the first imagery data,   creating a set of second annular data from an annulus of the second imagery data,   application of a frequency-domain transform on all sets of annular data to create a plurality of frequency-domain data sets,   calculating the complex conjugate of frequency-domain data,   multiplying the frequency-domain data sets together,   calculating an inverse frequency-domain transform of the resultant product of the multiplication, and   determining the translational offset between first and second imagery data by finding a maximal peak-to-side-lobe ratio for each result of the inverse transform step.   
     
     
         21 . The method of  claim 15 , wherein the image registration algorithm is a spatial-domain method comprising:
 selecting a two-dimensional patch of data derived from the second imagery data set,   conducting a convolutional step wherein the processor convolves the patch across data derived from the first imagery data set and at each position of the patch calculates a similarity metric, and   finding an extremum of all calculated similarity metrics.   
     
     
         22 . The method of  claim 21 , wherein the calculation of the similarity metric of the comprises a subtraction of data of two imagery data sets, and wherein the extremum is a minimum value of the similarity metric. 
     
     
         23 . The method of  claim 21 , wherein the calculation of the similarity metric of the comprises a multiplication of data of two imagery data sets, each datum having been adjusted by a factor derived from a sum of values within its respective imagery data set, and wherein that factor is also derived from the number of values in the sum, and wherein the extremum is a maximum value of the similarity metric. 
     
     
         24 . The method according to  claim 15 , further comprising:
 processing imagery data by an operator array comprising first or second order partial derivatives or absolute values thereof, or combination of first and second order partial derivatives or absolute values thereof, and   producing via the image registration function a set of two-dimensional coordinates, one for each partial derivative function in the operator array, and   wherein the method further comprises a step of logically analyzing these sets of two-dimensional coordinates to produce the translational offset.   
     
     
         25 . The method of  claim 15 , further comprising the determination of the point of impact of a projectile by:
 determining the difference between data elements derived from the first imagery data and those derived from the second imagery data,   filtering the output of the difference to determine a maximal value area.   
     
     
         26 . The method of  claim 25 , further comprising:
 detecting and removing line segments or edges residual of a difference between data elements derived from the first imagery data and those derived from the second imagery data.   
     
     
         27 . The digital weapon sight of  claim 15 , further comprising:
 validating a plurality of image registrations by comparing a plurality of translational offset outputs of image registration algorithms against each other.