An object detection method based on convolutional neural network adaptive background modeling
A convolutional neural network and self-adaptive background technology, applied in the field of background modeling, can solve problems such as no central pixel analysis, detection rate and accuracy rate not reached, feature extraction is not perfect, etc.
Active Publication Date: 2019-07-09
WUHAN UNIV OF TECH
View PDF6 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
[0005] (1) The method based on the region block is not perfect in the feature extraction of the region;
[0006] (2) There is no detailed analysis of the central pixel in the area block, and no good results have been achieved in terms of detection rate and correct rate
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
[0029] In order to facilitate those of ordinary skill in the art to understand and implement the present invention, the present invention will be described in further detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the implementation examples described here are only used to illustrate and explain the present invention, and are not intended to limit this invention.
[0030] please see figure 1 , a kind of object detection method based on convolutional neural network self-adaptive background modeling provided by the present invention, comprises the following steps:
[0031] Step 1: Use the first frame image to initialize the mixed Gaussian background model;
[0032] The mixed Gaussian model can be regarded as a function represented by multiple Gaussian distributions. The input of this function is the feature obtained by the area around the central pixel extracted by the convolutional neural network. The output of the func...
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 an adaptive background modeling object detection method based on a convolutional neural network. With respect to an existing background modeling method based on a region block, the method makes the following two improvements: 1) for the detect that description of discrete cosine transform coefficients serves as expression of regional characteristics in the existing background modeling method based on the region block, a region feature extraction method based on the convolutional neural network is provided to improve feature expression capability; and 2) an edge region central pixel type judgment method is provided, and by considering relation between an edge region center pixel and neighboring pixels thereof, whether the center pixel is a foreground pixel or a background pixel can be judged. Corresponding experimental verification shows that compared with the existing background modeling method based on the region block, the method, through two improvements, can bring a better effect.
Description
technical field [0001] The invention belongs to the technical field of background modeling, and relates to a background modeling object detection method based on area blocks, in particular to a convolutional neural network-based self-adaptive background modeling object detection method. Background technique [0002] In the field of background modeling technology, the process of the background modeling method based on the area block is generally divided into two parts: (1) divide each frame of image into several area blocks (small sample areas); (2) in the area block The background model is built on the level of the image (such a method can effectively use the context information of the image). [0003] In the research on the background modeling method based on the area block, H. Grabner proposed to use the advantage of online learning to train a series of the same classifiers for each background area block, and the sample area blocks with low confidence in the classifier are...
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): G06T7/262G06T7/277G06T7/254
CPCG06T7/254G06T7/262G06T7/277G06T2207/10016G06T2207/20021G06T2207/20081G06T2207/20084
Inventor 黄靖孙毅姜文周高景
Owner WUHAN UNIV OF TECH