US2023099984A1PendingUtilityA1
System and Method for Multimedia Analytic Processing and Display
Est. expiryApr 18, 2038(~11.7 yrs left)· nominal 20-yr term from priority
G06V 10/26G06T 2207/10116G06V 10/7715G06V 10/40G06V 40/1335G06V 10/462G06T 5/50
62
PatentIndex Score
0
Cited by
0
References
0
Claims
Abstract
The present disclosure includes systems and methods for multimedia image analytic including automated binarization, segmentation, and enhancement using bio-inspired based visual morphology schemes. The present disclosure further includes systems and methods for biometric multimedia content authentication using extracted geometric features and one or more of the binarization, segmentation, and enhancement methods.
Claims
exact text as granted — not AI-modified1 . A method of binarizing acquired input multimedia content, the method comprising the steps of:
a) receiving the input multimedia content; b) applying single window α-trim mean on one of the input multimedia content and a transformed grayscale channel of the input multimedia content; c) applying double window α-trim mean on one of the input multimedia content and the transformed grayscale channel of the input multimedia content; d) creating a visibility multimedia content using the single window α-trim mean and the double window α-trim mean; e) determining a local visual threshold using the visibility multimedia content and a visual morphological thresholding method; and f) generating a binarized multimedia content by applying the local visual threshold on the input multimedia content.
2 . The method of claim 1 and further comprising, prior to step b), performing a color space transformation on the input multimedia content and selecting a channel from the transformation to create the transformed grayscale channel of the input multimedia content.
3 . The method of claim 1 and further comprising, prior to step b), performing a color space transformation using an α-trim mean based principal component analysis (PCA) conversion on the input multimedia content, and selecting a channel from the transformation to create the transformed grayscale channel of the input multimedia content.
4 . The method of claim 1 , wherein applying the single window α-trim mean comprises:
dividing the input multimedia content into a plurality of content blocks;
performing a local α-trim mean on each of the plurality of content blocks; and
determining the single window α-trim mean, based on the local α-trim means.
5 . The method of claim 1 , wherein applying the double window α-trim mean comprises:
dividing the input multimedia content into a plurality of content blocks;
performing a local α-trim mean on each of the plurality of content blocks;
squaring a resulting α-trim mean for each of the plurality of content blocks;
dividing each squared result into a second plurality of content blocks;
performing the local α-trim mean on each of the second plurality of content blocks; and
determining a double window α-trim mean, based on the local α-trim means corresponding to the second plurality of content blocks.
6 . The method of claim 1 , wherein determining a local visual threshold comprises:
dividing the visibility multimedia content into a plurality of content blocks; determining a grey level density value for each of the plurality of content blocks; applying an optimization algorithm to each of the plurality of content blocks; and determining the local visual threshold.
7 . The method of claim 1 , wherein the input multimedia content is at least one forensic finger print image, and the binarized multimedia content is at least one binarized forensic finger print image.
8 . The method of claim 1 , wherein the input multimedia content is at least one microscopy image, and the binarized multimedia content is at least one binarized microscopy image.
9 . The method of claim 1 , wherein the input multimedia content is at least one 3D image, and the binarized multimedia content is at least one binarized 3D image.
10 . The method of claim 1 , wherein the input multimedia content is at least one X-ray image, and the binarized multimedia content is at least one binarized X-ray image.
11 . A method of segmenting an acquired multimedia content, the method comprising the steps of:
a) receiving the multimedia content; b) applying single window α-trim mean on one of the multimedia content and a transformed grayscale channel of the multimedia content; c) applying double window α-trim mean on one of the multimedia content and the transformed grayscale channel of the multimedia content; d) creating a visibility multimedia content using the single window α-trim mean and the double window α-trim mean; e) determining a local visual threshold using the visibility multimedia content and a visual morphological thresholding method; and f) generating a mask based on the local visual threshold; and g) generating a foreground multimedia content and a background multimedia content by applying the mask to the multimedia content.
12 . The method of claim 11 , wherein applying the single window α-trim mean comprises:
dividing the input multimedia content into a plurality of content blocks;
performing a local α-trim mean on each of the plurality of content blocks; and
determining the single window α-trim mean, based on the local α-trim means.
13 . A method of enhancing an acquired input multimedia content, the method comprising the steps of:
a) receiving the input multimedia content; b) determining a global visual threshold using the input multimedia content and a visual morphological thresholding method; c) creating a visual morphological enhanced multimedia content (VMEI/VMEV) by performing visual morphological equalization using the global visual threshold; d) creating a gamma corrected multimedia content by applying gamma correction to the VMEI/VMEV; and e) generating an enhanced multimedia content by fusing together the gamma corrected image with the input multimedia content.
14 . The method of claim 13 and further comprising determining a quality score of the VMEI/VMEV and repeating steps b) through c) until the quality score is above a threshold value.
15 . The method of claim 14 , wherein determining the quality score comprises:
applying single window α-trim mean on one of the input multimedia content and a transformed grayscale channel of the input multimedia content; applying double window α-trim mean on one of the multimedia content and a transformed grayscale channel of the multimedia content; computing a visibility multimedia content using the single window α-trim mean and the double window α-trim mean; and determining the quality score based on a mean intensity of the input multimedia content, the visibility multimedia content, and the VMEI/VMEV.
16 . A method of authenticating a biometric multimedia content, the method comprising the steps of:
a) receiving the biometric multimedia content; b) determining a global visual threshold using the biometric multimedia content and a visual morphological thresholding method; c) creating a visual morphological enhanced multimedia content (VMEI/VMEV) by performing visual morphological equalization using the global visual threshold; d) creating a gamma corrected multimedia content by applying gamma correction to the VMEI/VMEV; e) generating an enhanced biometric multimedia content by fusing together the gamma corrected multimedia content with the biometric multimedia content; f) creating a multimedia content template from the enhanced biometric multimedia content; g) obtaining a retrieved template; h) matching the multimedia content template with the retrieved template; and i) generating a report based on step h).
17 . The method of claim 16 and further comprising determining a quality score of the VMEI/VMEV and repeating steps b) through c) until the quality score is above a threshold value.
18 . The method of claim 16 , wherein step h) comprises:
performing feature extraction on the multimedia content template and the retrieved template to obtain multimedia content keypoints from the multimedia content template and template keypoints from the retrieved template; applying homography on the multimedia content template and the retrieved template; and calculating a match score based on a number of matching multimedia content keypoints and template keypoints.
19 . The method of claim 16 , wherein the biometric multimedia content is an image or video containing one of a fingerprint, a palm print, a footprint, a tongue image, a face, and an iris.
20 . The method of claim 16 , wherein the biometric multimedia content is a partial image or video, and further comprising employing a generative biometric completion technique comprising:
using a generator network which takes in random values to return a generated image, using a discriminator network that receives the generated multimedia content and defines an authenticity of the generated multimedia content, and outputting the generated image as a complete biometric multimedia content for use in steps b) and e).
21 . The method of claim 16 and further comprising obtaining images from multiple views or multiple sensors; performing feature detection; performing matching; eliminating irrelevant feature points detected; executing a RANSAC algorithm to obtain homography of the images; applying an alpha-trimmed correlation technique to find an optimal seam line to stitch; applying multi-level blending to construct a panoramic image that preserves biometric information; applying a segmentation technique to obtain relevant biometric traits; and performing biometric authentication based on the relevant biometric traits.Join the waitlist — get patent alerts
Track US2023099984A1 — get alerts on status changes and closely related new filings.
We store only your email — no account needed. See our privacy policy.