Face portrait generation method based on superpixel
A face portrait and super-pixel technology, applied in the field of image processing, can solve the problems of ignoring the structural information of the face image, losing the details of the pseudo-portrait, and not being able to effectively describe the local information of the face image, so as to achieve the effect of obvious details
Active Publication Date: 2018-05-04
XIDIAN UNIV
View PDF3 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
The disadvantage of this method is that the entire face image is represented as a vector, which cannot effectively describe the local information of the face image, and there is a loss of details in the process of portrait generation.
The disadvantage of this method is that each image block of the test photo is processed separately, ignoring the structural information of the face image itself, and the face structure is lost in the process of portrait generation.
The disadvantage of this method is that in the process of dividing the face image into blocks, the image is divided into rectangular blocks and the structural characteristics of the face image itself are ignored, and the generated pseudo-portrait has the loss of details.
Method used
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View moreImage
Smart Image Click on the blue labels to locate them in the text.
Smart ImageViewing Examples
Examples
Experimental program
Comparison scheme
Effect test
Embodiment Construction
[0027] refer to figure 1 , the steps that the present invention realizes are as follows:
[0028] Step 1, divide training portrait sample set, training photo sample set and test sample set.
[0029] Take M portraits from the portrait-photo pair set to form a training portrait sample set T R , and take out the training portrait sample set T R The portraits in the one-to-one correspondence of M photos constitute the training photo sample set T E , compose the remaining portrait-photo pairs into a test sample set, select a test photo S from the test sample set, 2≤M≤U-1, and U is the number of portrait-photo pairs in the portrait-photo pair set.
[0030] Step 2, divide the portraits in the training portrait sample set, the photos in the training photo sample set, and the test photos into superpixel blocks.
[0031] (2a) Divide the test photo S into superpixel blocks:
[0032] The division of the test photo S can be carried out using the existing superpixel algorithm. The supe...
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More PUM
Login to View More Abstract
The invention discloses a method for generating a human face portrait based on superpixels, which mainly solves the problem that the details of human face image segmentation are not obvious in the existing methods. Its realization steps are: (1) divide training portrait sample set, training photo sample set and test sample set; (2) carry out superpixel block division to the portrait in training portrait sample set, the photo in training photo sample set and test photo; 3) According to the divided super pixel blocks, the photo super pixel block set to be selected and the portrait super pixel block set to be selected are formed; (4) the portrait super pixel block weight set to be selected is calculated; The value set calculates the superpixel block set of the pseudo portrait; (6) generates the pseudo portrait according to the superpixel block set of the pseudo portrait. The present invention adopts superpixel block division in the block division process, and considers the structural information of the face image itself, so that the details of the generated face portrait are obvious, and can be used for face retrieval and recognition in the field of public security.
Description
technical field [0001] The invention belongs to the technical field of image processing, and in particular relates to a method for generating a face portrait, which can be used for face retrieval and recognition in the field of public security. Background technique [0002] In public security management, face-based identification is one of the most convenient and effective authentication technologies. For example, in criminal investigations, when it is difficult to obtain photos of the suspect, the forensic doctor can draw a portrait of the suspect based on the description of the witnesses. Due to the different imaging principles of photos and portraits, there are great differences in shape and texture, which brings many difficulties to face recognition based on portraits. Face portrait generation technology is to convert the photos in the police face database into fake portraits through image processing technology, and then the suspect's portrait can be identified in the f...
Claims
the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More Application Information
Patent Timeline
Login to View More Patent Type & Authority Patents(China)
IPC IPC(8): G06K9/62
CPCG06F18/214
Inventor 高新波彭春蕾王楠楠李洁朱明瑞于昕晔张声传张铭津孙雷雨张宇航
Owner XIDIAN UNIV