Human-Shape Image Segmentation Method

• US 20170200274A1
• Filed: 05/23/2014
• Published: 07/13/2017
• Est. Priority Date: 05/23/2014
• Status: Active Grant

• Current Assignee: Watrix Technology
• Original Assignee: Watrix Technology
• Inventors: WANG, Liang, TAN, Tieniu, HUANG, Yongzhen, WU, Zifeng
• Granted Patent: US 10,096,121 B2
• Time in Patent Office: Days
• CPC Class Codes: G06F 18/2415 based on parametric or probabilistic models, e.g. based on likelihood ratio or false acceptance rate versus a false rejection rate
• CPC Class Codes: G06N 3/045 Combinations of networks
• CPC Class Codes: G06N 3/08 Learning methods
• CPC Class Codes: G06N 3/084 Backpropagation, e.g. using gradient descent
• CPC Class Codes: G06N 7/01 Probabilistic graphical models, e.g. probabilistic networks
• CPC Class Codes: G06T 2207/20016 Hierarchical, coarse-to-fine, multiscale or multiresolution image processing; Pyramid transform
• CPC Class Codes: G06T 2207/20021 Dividing image into blocks, subimages or windows
• CPC Class Codes: G06T 2207/20076 Probabilistic image processing
• CPC Class Codes: G06T 2207/20081 Training; Learning
• CPC Class Codes: G06T 2207/20084 Artificial neural networks [ANN]
• CPC Class Codes: G06T 2207/30196 Human being; Person
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06T 7/143 involving probabilistic approaches, e.g. Markov random field [MRF] modelling
• CPC Class Codes: G06V 10/454 Integrating the filters into a hierarchical structure, e.g. convolutional neural networks [CNN]
• CPC Class Codes: G06V 10/50 by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
• CPC Class Codes: G06V 30/194 References adjustable by an adaptive method, e.g. learning
• CPC Class Codes: G06V 30/2504 Coarse or fine approaches, e.g. resolution of ambiguities or multiscale approaches
• CPC Class Codes: G06V 40/103 Static body considered as a whole, e.g. static pedestrian or occupant recognition

Abstract

A human-shape image segmentation method comprising: extracting multi-scale context information for all first pixel points for training a human-shape image; sending image blocks of all scales of all the first pixel points into a same convolution neural network to form a multi-channel convolutional neural network group, wherein each channel corresponds to image blocks of one scale; training the neural network group using a back propagation algorithm to obtain human-shape image segmentation training model data; extracting multi-scale context information for all second pixels points for testing the human-shape image; sending image blocks of different scales of each of the second pixel points into a neural network channel corresponding to the human-shape image segmentation training model, wherein if said first probability is larger than said second probability, the second pixel points belong to the human-shape region, otherwise, the second pixel points are outside of the human-shape region. The human-shape image segmentation method is fast in image segmentation speed and high in accuracy.

Claims

1. A human-shape image segmentation method, characterized by comprising:

step S1;

extracting multi-scale context information for all first pixel points for training a human-shape image;

step S2;

sending image blocks of all scales of all the first pixel points into a same convolution neural network to form a multi-channel convolutional neural network group, wherein each channel corresponds to image blocks of one scale;

step S3;

training the neural network group using a back propagation algorithm to obtain human-shape image segmentation training model data;

step S4;

extracting multi-scale context information for all second pixels points for testing the human-shape image;

step S5;

sending image blocks of different scales of each of the second pixel points into a neural network channel corresponding to the human-shape image segmentation training model, wherein all of said neural network channels are merged together in a full-connected layer, a first value representing a first probability of said second pixel points belonging to the human-shape region is output at a first node of the last layer of the full-connected layer, and a second value representing a probability of said second pixel points being outside of the human-shape region is output at a second node of the last layer of the full-connected layer;

if said first probability is larger than said second probability, the second pixel points belong to the human-shape region, otherwise, the second pixel points are outside of the human-shape region.

2. The method according to claim 1, characterized in that said step S1 specifically includes:

for all first pixel points inside and outside of the human-shape region for training the human-shape image, extracting image blocks of different scales by using the first pixel point as the center.

3. The method according to claim 1, characterized in that in said step S2, the multi-channel convolutional neural network group are merged together in the full-connected layer;

the last layer of the full-connected layer is an output layer and comprises a first node and a second node, and output of all pixel points inside the human-shape region corresponds to said first node, and output of all pixel points outside the human-shape region corresponds to said second node.

4. The method according to claim 1, characterized in that said step S4 specifically includes:

for all second pixel points for testing the human-shape image, extracting image blocks of different scales by using said second pixel point as the center.