Meeting the needs of the imaging research community
                                                                         Haze Remover


Aerial images often contain haze, hiding critical information about the scene. Image Registration and Fusion Systems has developed a
haze remover software that can interactively or automatically remove haze from an image. The interactive method allows the user to see
local details in an image through a circular window as shown in image (a) below. The automatic method removes haze over the entire
image automatically and displays the enhanced image, as shown in (b).


















The haze removed image can be saved, if required. The software allows removing haze from a sequence of images and save the enhanced
images.















                                                  Multi-exposure and Multi-focus Image Fusion

It is impossible to capture an image of a 3-D scene that is well-exposed and well focused everywhere. Lighting across the scene can vary
and  an image appears sharp only at scene points appearing at or near the focus plane. To create an image that is well exposed and well
focused everywhere, a number of images are obtained at different focus levels and different exposure level and the images are combined. A
software developed by Image Registration and Fusion Systems can combine (fuse) such images into an image that is well-exposed and
well-focused everywhere. Image fusion is achieved by maximizing image entropy. This ensures that the fused image contains more
information about the scene than any other fused image obtained from the same images. An example demonstrating fusion of
multi-exposure images by this software is given below. Five images obtained at different exposure levels are combined into a single
well-exposed image.























Fusion of multi-focus images by this software is demonstrated below. Image (c) is obtained by fusing images (a) and (b).















Image fusion requires that the provided images be registered. A version of the software that registers the images before fusing them is
available. An example of this software is given below. Images (a) - (d) show images obtained at different exposure levels. These images
are slightly shifter with respect to each other. Image (e) shows fusion of images (a) - (d) without image registration and image (f) shows
fusion of the same images after registration.





























The fused image can be saved for further use.











































For information about Image Registration and Fusion Systems' products and services, please contact: cs@imgfsr.com
Image Registration and Fusion Systems, Beavercreek, Ohio, U.S.A., Copyright 2016
                                         (a)                                                                                                 (b)
                                                     Detecting Line Landmarks in an Image

Lines are geometric features that are not only useful in image registration, but they are also useful in various image analysis applications.
Line landmarks are abundant in images of man-made scenes. A software tool developed by Image Registration and Fusion Systems
detects lines in an image by various methods, including 1) Hough transform, 2) clustering, 3) contour tracing, 4) curve fitting, and 5) region
subdivision. Examples of line detection by (a) contour tracing and (b) curve fitting are given below.


























In addition to displaying the detected lines, the software allows saving the coordinates of the endpoints of the lines for later use.
Software Products
Terms of Sale
  1. All sales are final. To ensure that a software meets your needs, before purchase, please contact: cs@imgfsr.com
  2. A license includes technical support for 3 months and upgrades for 1 year.
  3. Each license comes with a User's Guide, describing usage of the software.
  4. Software licenses are for Windows PC only. Each software is standalone and complete.
  5. Special orders to add new functions to an existing software will be considered.
  6. A single-user license is for use by a single user. For site licenses, please contact: cs@imgfsr.com
  7. For ordering and delivery information, please contact: cs@imgfsr.com
  8. For technical support, please contact: ts@imgfsr.com
                                                  (a)                                                                                                                           (b)
                                                Detecting an Image's Dominant Orientation


This software finds the dominant orientation of an image and prints it to the screen. It also draws a line through the image center showing
the detected dominant orientation. An example where the detected dominant orientation is 90 degrees is given below..































Given two images of a scene, this software will also rotate the second image to the orientation of  the first. For example, given images (a)
and (b) below, the software rotates image (b) to the orientation of image (a), as shown in (c).
                             (a)                                                                               (b)                                                                               (c)
                                              Detecting Point Landmarks in an Image


Point landmarks identify locally unique neighborhoods in an image. Depending on how uniqueness is defined, different point landmarks
can be detected. This software detects various types of point landmarks, including 1) local extrema of image central moments, 2) local
extrema of image Laplacian of Gaussian, 3) local maxima of image entropy, 4) local maxima of image invariant moments of order 2, 5)
Harris corners, and 6) Hessian corners. The software allows specifying the desired number of point landmarks. An example detecting
100 point landmarks in an image by the Harris corner detector is given below.































The software allows saving the coordinates of the detected points for later use.
For questions or additional information, please contact: cs@imgfsr.com
To obtain a single-user license
for this software, follow this link =>  
For questions or additional information, please contact: cs@imgfsr.com
To obtain a single-user license
for this software, follow this link =>
                                                    Finding Corresponding Points in Images


Establishing correspondence between point landmarks in two images of a scene makes it possible to find the geometric relation between
the images. Corresponding points can be used to track moving objects, find scene geometric changes, and recognize objects. A software
developed by Image Registration and Fusion Systems finds corresponding points in images of a scene. An example of point
correspondence by this software is given below. 57 corresponding points are obtained in images (a) and (b). Corresponding points are
shown with the same color in the images. The user by controlling the number of points detected in each image controls the number of  
corresponding points found in the images.


















The software uses the coordinates of corresponding points to transform the second image to the geometry of the first. The images are then
spatially aligned, as shown in (c). The coordinates of corresponding points as well as the transformed image can be saved for later use..
                           (a)                                                                               (b)                                                                               (c)
For questions or additional information, please contact: cs@imgfsr.com
To obtain a single-user license
for this software, follow this link =>
                                                     Finding Corresponding Lines in Images


A software developed by Image Registration and Fusion Systems finds corresponding points in two images of a scene by first finding a set
of lines in each image and then finding correspondence between the lines. An example of line correspondence by this software is given
below. 115 corresponding lines are found between images (a) and (b). Corresponding lines are shown with the same color.


















The software uses information about the line correspondences to transform the second image to the geometry of the first and spatially align
(register) the images, as shown in (c). The coordinates of the endpoints of corresponding lines as well as the transformed image can be
saved for later use.
                         (a)                                                                               (b)                                                                               (c)
For questions or additional information, please contact: cs@imgfsr.com
To obtain a single-user license
for this software, follow this link =>
                                                                  Nonrigid Image Registration


Analysis of two or more images of a scene often requires registration of the images. If the images to be registered have local geometric
differences, there is a need to find correspondence between point landmarks in images and from the correspondences estimate the
parameters of an elastic transformation to transformed one image to the geometry of the other so the images can be spatially aligned and
compared. An example of nonrigid image registration is given below. Images (a) and (b) show aerial images of a 3-D scene captured from
different views. 118 corresponding points are obtained in the images. Corresponding points are shown with the same color. The
coordinates of corresponding points are used to determine the parameters of an elastic transformation function to transform image (b) to
the geometry of image (a). Overlaying of image (a) and image  (b) after the elastic transformation is shown in (c).





































The software allows saving the coordinates of corresponding points in the images as well as the transformed image for later use. The
means to evaluate the registration quality is also provided.
                                                        (a)                                                                                                                         (b)
                                                      (c)
For questions or additional information, please contact: cs@imgfsr.com
To obtain a single-user license
for this software, follow this link =>
                                                                 Volume Image Registration


A volume image registration software developed by Image Registration and Fusion Systems automatically registers two tomographic image
sequences. The images can be in the same modality or in different modalities, and the images can be related by a rigid or a nonrigid
transformation. The software finds sufficient corresponding point landmarks in the images to determine the parameters of a transformation
function to  resample one image to the geometry of the other. An example demonstrating registration of MR and PET brain images is given
below. These images are courtesy of NIH. The images before registration are shown in (a) and the images after registration are shown in
(b). A volumetric is displayed by its orthogonal cross-sections. The software is fully automatic and takes only a few minutes to register the
images.

























Tools are provided to visually and quantitatively evaluate the registration accuracy. The software also allows saving the resampled image for
later use.
                                                             (a)                                                                                                                   (b)
For additional information or to obtain a single-user license for this software, please contact: cs@imgfsr.com
Price
For questions or additional information, please contact: cs@imgfsr.com
To obtain a single-user license
for this haze remover, follow this link =>
For questions or additional information, please contact: cs@imgfsr.com
To obtain a single-user license
for this software, follow this link =>
                                   (a)                                                                             (b)                                                                               (c)
                     (a)                                                           (b)                                                             (c)                                                           (d)
                                                            (e)                                                                                                                             (f)    
For questions or additional information, please contact: cs@imgfsr.com
To obtain a single-user license for the multi-exposure
      and multi-focus fusion software, follow this link =>
To obtain a single-user license for the multi-exposure
        and multi-focus fusion software, including the
                       registration software,  follow this link =>
Price
Price
Price
Price
Price
Price
Price
Price