The invention relates to a method of inserting an insertion object into a sequence of images. The insertion object may be an image, a video, or a three-dimensional model, which could possibly be animated. Particularly, but not exclusively, the invention relates to the insertion of advertisement images into video, such as videos of sporting events. A method comprises capturing a sequence of images, the sequence of images comprising in order a first image, a second image, and a third image; estimating a first homographic transform from the first image to the third image; deriving a second homographic transform from the first image to the second image based on the first homographic transform; transforming the insertion object using the first homographic transformation to form a first warped insertion image, and inserting the first warped insertion image into the third image of the sequence of images; and transforming the insertion object using the second homographic transformation to form a second warped insertion image, and inserting the second warped insertion image into the second image of the sequence of images.

A method for identifying authenticity of an object, the method includes maintaining, in an identification server system, a reference image of an original object, the reference image and provided to represent all equivalent original objects, receiving, in the identification server system, one or more input images of the object to be identified, and generating, by the identification server system, a target image from the one or more input images. The method further includes aligning, by the identification server system, the target image with the reference image and analysing, by the identification server system, the target image in relation to the aligned reference image for identifying authenticity of the object.

A two-stage medical image registration method based on progressive images (PIs) to solve the technical problem of low registration accuracy of traditional image registration methods includes: merging a reference image with a floating image to generate multiple intermediate PIs; registering, by a speeded-up robust features (SURF) algorithm and an affine transformation, the floating image with the intermediate PIs to acquire coarse registration results; registering, by the SURF algorithm and the affine transformation, the reference image with the coarse registration results to acquire fine registration results; and comparing the fine registration results of the intermediate PIs, which are acquired by iteration, and selecting an optimal registration result as a final registration image. The method can achieve multimodal registration for brain imaging with MI, NCC, MSD, and NMI superior to those of the existing registration algorithms. The method effectively improves the registration accuracy through the progressive medical image registration strategy.

A medical image processing device a reference image that is a medical image with which boundary line information related to a boundary line that is a boundary between an abnormal region and a normal region and landmark information related to a landmark that is a characteristic structure of the subject are associated and a captured image that is the medical image captured in real time, detects the landmark from the captured image, calculates a ratio of match between the landmark included in the reference image and the landmark included in the captured image, estimates a correspondence relationship between the reference image and the captured image on the basis of the ratio of match and information regarding the landmarks included in the reference image and the captured image, and generates a superimposition image in which the boundary line associated with the reference image is superimposed on the captured image on the basis of the correspondence relationship.

An image stitching method is proposed to include: A) acquiring a plurality of segment images for a target scene, each of the segment images containing a part of a target scene; B) for two adjacent segment images, which are two of the segment images that have overlapping fields of view, comparing the two adjacent segment images to determine a stitching position for the two adjacent segment images from a common part of the overlapping fields of view; and C) stitching the two adjacent images together based on the stitching position thus determined.

Systems and methods for iteratively computing an image registration or an image segmentation are driven by an optimization function that includes a similarity measure component whose effect on the iterative computations is relatively mitigated based on a monitoring of volume changes of volume elements at image locations during the iterations. A system and a related method quantify a registration error by applying a series of edge detectors to input images and combining related filter responses into a combined response. The series of filters are parameterized with a filter parameter. An extremal value of the combined response is then found and a filter parameter associated with said extremal value is then returned as output. This filter parameter relates to a registration error at a given image location.

A method and device are for automatic determination of the change of a hollow organ. The method includes providing a first medical image of the organ recorded at a first time; computing a first representation of the organ in the first image; computing a first reference-line of the organ based on the first representation and providing a second medical image of the organ recorded at a second point. The method further includes computing a second representation of the organ in the second image; computing a second reference-line of the organ based on the second representation of the organ; registering of the first and second reference-line to obtain at least one of matched representations of the organ and features derived from the matched representations of the organs; and comparing at least one of the matched representations of the organs and the features derived from the matched representations of the organ.

Disclosed are a video stitching method and a video stitching device. The video stitching method is applicable for stitching a first video and a second video, and includes: performing feature extraction, feature matching and screening on a first target frame of the first video and a second target frame of the second video, so as to obtain a first feature point pair set; performing forward tracking on the first target frame and the second target frame, so as to obtain a second feature point pair set; performing backward tracking on the first target frame and the second target frame, so as to obtain a third feature point pair set; and calculating a geometric transformation relationship between the first target frame and the second target frame according to a union of the first feature point pair set, the second feature point pair set and the third feature point pair set.

Discussed herein are devices, systems, and methods for multi-image ground control point (GCP) determination. A method can include extracting, from a first image including image data of a first geographical region, a first image template, the first image template including a contiguous subset of pixels from the first image and a first pixel of the first image indicated by the GCP, predicting a first pixel location of the GCP in a second image, the second image including image data of a second geographical overlapping with the first geographical region, extracting, from the second image, a second image template, the second image template including a contiguous subset of pixels from the second image and a second pixel corresponding to the pixel location, identifying a second pixel of the second image corresponding to a highest correlation score, and adding a second pixel location of the identified pixel to the GCP.

A method for training a machine learning model to generate a predicted measured image, the method including obtaining (a) an input target image associated with a reference design pattern, and (b) a reference measured image associated with a specified design pattern printed on a substrate, wherein the input target image and the reference measured image are non-aligned images; and training, by a hardware computer system and using the input target image, the machine learning model to generate a predicted measured image.