Patent
Fingerprint image completion simulation method and system
TLDR
Wang et al. as mentioned in this paper proposed a fingerprint image completion simulation method and system, which comprises the steps that multiple fingerprints are collected, and incomplete fingerprint images and complete fingerprint images are judged and distinguished; convolution network responses are extracted from the incomplete fingerprints, and superpixel segmentation is performed on the incomplete images; CNN response aggregation features of superpixels are extracted according to the CNN responses and a superpixel image segmentation result; image blocks with structural characteristics in complete fingerprint image information are matched to obtain an affine mapping relation between similar regions.Abstract:
The invention discloses a fingerprint image completion simulation method and system The method comprises the steps that multiple fingerprint images are collected, and incomplete fingerprint images and complete fingerprint images are judged and distinguished; convolution network responses are extracted from the incomplete fingerprint images, and superpixel segmentation is performed on the incomplete fingerprint images; convolution network response aggregation features of superpixels are extracted according to the convolution network responses and a superpixel segmentation result; a superpixel structure adjacent to an unknown region is utilized to perform similar structure graph matching on a superpixel adjacent graph, and a middle-level structure region similar to the unknown region is obtained; image blocks with structural characteristics in complete fingerprint image information are matched to obtain an affine mapping relation between similar regions; and optimal image blocks used for completing the unknown region are optimized and acquired in combination with the affine mapping relation and a Markov random field energy minimization model, and image fusion is performed to generate a completed fingerprint image The fingerprint image completion simulation method and system aim to solve the technical problem that a complete fingerprint image cannot be obtained when a user fingerprint is incompleteread more
Citations
More filters
Patent
Convolutional neural network based power defect image recognizing method
TL;DR: In this article, a convolutional neural network based power defect image recognizing method was proposed, which was characterized in that aneural network structure was designed; an image detecting and property description algorithm in the special field was built by the network optimizing method, and whether the image has defect and the type of the defect can be directly obtained through the neural network; and the area with the defect could be output at the same time.
Patent
Control method, control device and mobile terminal
TL;DR: In this paper, a control method, a control device, a mobile terminal and a computer readable storage medium are presented for controlling a mobile device with the same fingerprint input at different screen positions.
Patent
Security inspection intelligent auxiliary image judgment method and system and system control method
TL;DR: In this article, a security inspection intelligent auxiliary image judgment method and system and a system control method is proposed, and the method comprises the steps of carrying out the classification of an original image obtained through the multi-angle scanning of an X-ray machine based on an image color threshold value, carrying out segmentation of the classified original image based on edge extraction algorithm, and achieving the preliminary extraction of all articles in the image; on the basis of the adversarial generation network in the deep learning algorithm, complementing the contours of the images of all the objects after segmentation
Patent
Gesture recognition method and device and computer readable storage medium
TL;DR: In this article, a gesture recognition method and device and a computer readable storage medium is presented, and the method comprises the steps: carrying out human body key point detection on a target image, and obtaining at least one human key point; under the condition that the at least 1 human body point comprises a wrist point, determining a recognition area, and comprising a position in the target image and the distance between the position and the wrist point meets a preset condition.
References
More filters
Patent
Application of Z-Webs and Z-factors to Analytics, Search Engine, Learning, Recognition, Natural Language, and Other Utilities
Saied Tadayon,Bijan Tadayon +1 more
TL;DR: In this paper, the authors introduce Z-webs, including Z-factors and Z-nodes, for the understanding of relationships between objects, subjects, abstract ideas, concepts, or the like, including face, car, images, people, emotions, mood, text, natural language, voice, music, video, locations, formulas, facts, historical data, landmarks, personalities, ownership, family, friends, love, happiness, social behavior, voting behavior, and the like.
Patent
Fingerprint identification method and device
TL;DR: In this article, the authors proposed a fingerprint identification method and device, which comprises the following steps: continuously carrying out a fingerprint image collection on a target finger at least twice, and obtaining at least two copies of initial fingerprint image data of the target finger; and calculating the mean value of pixel values of corresponding pixel points in each copy of the initial image data, mapping the mean values into the corresponding pixels, and carrying out fingerprint identification according to the target image data.
Proceedings ArticleDOI
Fingerprint image enhancement using CNN Gabor-Type filters
Ertugrul Saatci,Vedat Tavsanoglu +1 more
TL;DR: This paper proposes fingerprint image enhancement based on CNN Gabor-type filters that can effectively enhance fingerprint ridges by selecting a directional filter with correct filter parameters to match ridge features at each point.
Patent
Method of quickly segmenting moving target in non-restrictive scene based on full convolution network
TL;DR: In this paper, a method of segmenting a moving target in a non-restrictive scene based on a full convolution network is proposed, which belongs to the technical field of video object segmentation, and the method comprises steps: firstly, framing is carried out on the video, and a result after framing is used for making a Ground Truth set S for a sample image; a deep feature estimator for an image foreground target is acquired, target maximum intra-class likelihood mapping information in all frames is obtained, and initial prediction on the foreground and the background in
Patent
Road detection based on superpixels and convolution neural network
TL;DR: In this paper, a road detection method based on superpixels and a convolution neural network is proposed, and the method comprises the steps: carrying out the image preprocessing through a simple linear iteration cluster algorithm, and segmenting an image into super pixels which are uniform in size.