Compute the DEM covariance for the given problem, and also the albedo covariance, if --float-albedo is on. If --save-covariances is on, also save cova

Calculate the DEM slope and aspect in degrees. Use the same logic as gdaldem.

Calculate world_to_cam transforms from their representation in a vector, rather than using reference cameras, extrinsics and timestamp interpolation.

Compute a matrix transform between images, searching for IP in the specified regions.

Compute the ECEF transform (around planet center) given the interest point matches

Compute the lon-lat box and height range from the DEM

Calculate the maximum displacement from the source points (after any initial transform is applied to them) to the source points after alignment with t

Given a calendar time, find the midnight time. Just put zeros for hours, minutes, and seconds. An input time looks like: 2022-04-13T22:46:31.4540000

Given the transforms from each camera to the world and their timestamps, find an initial guess for the relationship among the sensors on the rig. Note

Compute output georeference to use

Compute the translation vector from the source points (before any initial alignment applied to them), and the source points after alignment.

Find cam type based on cam name

The camera will be constrained by the ground, but not by the roll/pitch/yaw, then the orientation will change along the trajectory. Then adjust the al

Generate a camera model from a single ortho file

Return the exact camera pose

See --extrinsics.

Compute centerline weights using plateaued weight function (not normalized, unlike VW's version). Distinguishes interior holes from border pixels and

Replace .<extension> with <suffix>

Change the color of given layer of polygons

Return true if all the images have been converted from jpeg.

Check if an HTML file has the 404 Not Found message.

Return true if the given IceBrige folder URL is valid. This functionality, if done right, times out, likely to pecularities of the NSIDC serve

Get a count of the label files produced in a frame range.

See if we need to estimate the proj win of the source points based on reference points and max disp. This is needed for COPC files when this win is no

Verify that we have all the tools we will be calling during the script.

Verify that we have all the tools we will be calling during the script.

Throws if the input point clouds do not have the same number of channels. - Returns the number of channels.

Convert any LAS, CSV, PCD, or unorganized TIF files to ASP tif files. We do some binning to make the spatial data more localized, to improve performan

Clean up all non-output files to conserve space. Setting smallFiles will remove additional low size files.

Recursive function to eliminate extraneous nodes created by our parsing technique

Parse format: "focal_length optical_center other_intrinsics" Applies to all sensors and when not optimizing intrinsics per sensor.

Estimate the height error at each DEM pixel based on estimates for each image

Given a cnet between a set of left images, and disparities from left to right images, add more interest points to the cnet based on the disparities.

Given a vector, find a rotation matrix that rotates the vector to the y-axis.

Compute the angle from the top left to the top right of an image. Positive rotation = clockwise.

Compute block size for rasterization based on bias and user options. Block size is set to next power of 2 >= 4*bias, with minimum of 256.

See how well the optimized model fits the ground points

Compute weights for a DEM in the mosaic. This applies a pipeline of operations: grassfire distance, erosion/centerline, clamping, blurring, power func

Compute the Nuth offset. By now the DEMs have been regrided to the same resolution and use the same projection. All data is kept as float, but the cal

Compute a rigid transform between n point correspondences. There exists another version of this using vw matrices in VisionWorkbench called find_3D_tr

Compute the bias for reading from images beyond tile boundaries to avoid tiling artifacts. This accounts for erosion, blurring, hole filling, etc.

Compute the transform from the B map to the A map by finding the median transform based on the shared images

Compute the positions of each image relative to the first image. - The top left corner of the first image is coordinate 0,0 in the output image.

Compute ip between L.tif and R.tif produced by stereo.

Compute a matrix transform between images, searching for IP in the specified regions.

TODO: Pass in image ref instead of paths? Compute the transform from image1 to image2 (the top left corner of image1 is (0,0))

Consolidate statistics into a single file

Contour the current image

TODO(oalexan1): The 0.5 bias may be the wrong thing to do. Need to test more overlaying an image which is above threshold in a rectangular region and

If the text is a number returns float(text), otherwise just text

Given the json metadata, generate a .tsai camera file

Populate from an AdjustedCameraModel

Populate from CSM. Since with CSM we apply adjustments to existing cameras, these start as 0.

Populate from OpticalBarModel

Populate from a PinholeModel

This is a bugfix for when points far from the field of view project incorrectly into the camera.

Wrap a widget in a horizontal layout with a colorbar on the right. Returns the wrapper widget, or the original widget if bounds are invalid. The color

Create a DEM from a pair of images

Create the image masks for the left and right images

Create pixel samples. Make sure to sample the pixel at (width - 1, height - 1).

Create structures for pixels, xyz, and weights, to be used in optimization. Later there will be another pass to add weights for the anchor points. Her

Create a copy of the input camera model which accounts for a 90 degree rotation of the corresponding image.

Create subsampled images and masks

Create symlinks to the input images for skip_image_normalization mode. The symlinks are relative to the output directory.

Take input cameras and corrections in param_storage, and create new cameras incorporating the corrections.

Compute a ground plane, scale the orbital camera positions relative to the plane, then plot the cameras and the ground plane.

Crop out a section of an image

Read the CSM state (a string) from a cube file. Throw an exception if missing. Careful not to copy junk from the blob.

Return either xyz or a projected point. Note that the flag return_point_height is not necessarily respected.

Return the OpenCV matrix type as a string

Check that two georefs have the same projection. Compare PROJ4 strings and dynamic datum properties. WKT strings can differ in naming conventions whil