Recent Patents in Signal Processing (September 2016) – Image registration

For our August 2016 issue, we cover recent patents dealing with hardware and algorithmic issues of image registration. The section below covers patents granted recently for rigid and elastic image registration, patch matching, image alignment under low contrast or illumination, and specific devices constructed for the task.

The disclosure presented in Patent no. 9,418,468 introduces a method and device for elastic registration of a two-dimensional source digital image of an object of interest. A slice of a three dimensional target volume of the object of interest is provided which defines a Markov Random Field framework having at least one undirected pairwise graph superimposed a two-dimensional image domain having at least a set of regular vertices and at least a set of edges, and defines a grid of control points, each control point corresponding to a vertex of the set of regular vertices, and a neighborhood system of edges associated with vertices. The method also defines a set of multi-dimensional labels of a discrete space to apply a displacement to each control point, the control point displacement defining a transformation adapted to obtain a plane slice of the target volume and an in-plane deformation transformation of the source digital image.

Various embodiments of patent no. 9,414,037 provide a method of enhancing images in low lighting conditions wherein a color image is captured with a color camera, a first monochromatic image is captured with a first monochromatic camera, and a second monochromatic image is captured with a second monochromatic camera. Image registration is performed to align the first and second monochromatic images and the luminance information from the registered monochromatic images is fused with chrominance information from the color image.

As descibed in patent no. 9,411,253, the rotation speed of an intermediary transfer belt when it is brand-new is stored as a reference rotation speed. Registration patches Rk to Ry are formed to calculate displacement amounts, and based on the rotation speed difference between the rotation speed of the intermediary transfer belt  and the reference rotation speed, the calculated displacement amounts are corrected into displacement amounts expected when the secondary transfer roller is contacted with the intermediary transfer belt.  Next, based on the corrected displacement amounts, the positions at which exposing devices 60K to 60Y form electrostatic latent images are corrected. Thus, registration correction can be performed accurately without requiring too much time or toner, with consideration given to the difference in the rotation speed of the intermediary transfer belt depending on whether the secondary transfer roller is contacted with or released from the intermediary transfer belt.

In patent no. 9,406,133 a method for image registration includes the steps of: providing a computer and an imager; acquiring and processing a reference image frame including a plurality of reference strips; acquiring and correcting a target frame by the steps of: acquiring one or more target frame strips of the target frame; comparing the one or more target frame strips in real-time to one or more corresponding strips of the reference frame based on a calculated registration of a specific target frame strip from a previous target frame; correcting a registration of the one or more target frame strips; repeating steps until a last strip of the target frame; repeating the steps of acquiring and correcting a target frame until a last target frame is acquired; and saving processed image based on a plurality of corrected target frames to a non-volatile memory. A system to perform the method is also described.

In the invention no. 9,384,552 systems, methods, and computer readable media to improve image stabilization operations are described. A novel approach to pixel-based registration of non-reference images to a reference frame in a set of commonly captured images is disclosed which makes use of pyramid decomposition to more efficiently detect corners. The disclosed pixel-based registration operation may also be combined with motion sensor data-based registration approaches to register non-reference images with respect to the reference frame. When the registered non-reference images are combined with the pre-selected reference image, the resulting image is a quality stabilized image.

A system and method is presented in patent no. 9,383,210 for improved image navigation and registration on a low cost remote sensing satellite based on reference image data received from an exquisite system.

In patent no. 9,361,692 an image registration device is provided including a first feature vector magnitude calculating unit calculating magnitudes of feature vectors corresponding to any one first feature point among feature points of a reference image to create a first magnitude value, a second feature vector magnitude calculating unit calculating magnitudes of feature vectors corresponding to any one second feature point among feature points of a target image to create a second magnitude value, a magnitude difference calculating unit receiving the first and second magnitude values and calculating a difference between the received first and second magnitude values to create a third magnitude value, a first threshold value creating unit creating a first threshold value on the basis of the first magnitude value and a magnitude ratio, and a magnitude difference determining unit receiving the third magnitude value and the first threshold value, and determining a magnitude difference.

If you have an interesting patent to share when we next feature patents related to image registration, or if you are especially interested in a signal processing research field that you would want to be highlighted in this section, please send email to Csaba Benedek (benedek.csaba AT sztaki DOT mta DOT hu).

References

Number: 9,418,468
Title: Method and device for elastic registration between a two-dimensional digital image and a slice of a three-dimensional volume with overlapping content
Inventors: Paragios; Nikos (Paris, FR), Ferrante; Enzo (Paris, FR), Marini Silva; Rafael (Paris, FR)
Issued: August 16, 2016
Assignee: Centralesupelec (Gif sur Yvette, FR)

Number: 9,414,037
Title: Low light image registration
Inventors: Solh; Mashhour (San Jose, CA)
Issued: August 9, 2016
Assignee: Amazon Technologies, Inc. (Reno, NV)

Number: 9,411,253
Title: Image forming apparatus forming registration patches for detection of displacement of toner images
Inventors: Naruse; Kunihiko (Toyokawa, JP), Kawahara; Morihisa (Toyohashi, JP)
Issued: August 9, 2016
Assignee: Konica Minolta, Inc. (Tokyo, JP)

Number: 9,406,133
Title: System and method for real-time image registration
Inventors:  Yang; Qiang (Rochester, NY)
Issued: August 2, 2016
Assignee: University of Rochester (Rochester, NY)

Number: 9,384,552
Title: Image registration methods for still image stabilization
Inventors: Zhou; Jianping (Fremont, CA), Hubel; Paul (Mountain View, CA), Tico; Marius (Mountain View, CA), Nariani Schulze; Anita (Los Altos, CA), Toft; Rolf (Palo Alto, CA)
Issued: July 5, 2016
Assignee: Apple Inc. (Cupertino, CA)

Number: 9,383,210
Title: Image navigation and registration (INR) transfer from exquisite systems to hosted space payloads
Inventors: Carr; James L. (Washington, DC)
Issued: July 5, 2016

Number: 9,361,692
Title Image registration device and operation method of the same
Inventors:  Suk; Jung Hee (Daejeon, KR), Lyuh; Chun-Gi (Daejeon, KR), Chun; Ik Jae (Daejeon, KR), Roh; Tae Moon (Daejeon, KR)
Issued: June 7, 2016
Assignee: Electronics and Telecommunications Research Institute (Daejeon, KR)