Pedestrian detection method and system based on deep neural network, equipment and medium
What is Al technical title?
Al technical title is built by PatSnap Al team. It summarizes the technical point description of the patent document.
A deep neural network, pedestrian detection technology, applied in the field of pedestrian detection based on deep neural network
Pending Publication Date: 2021-06-08
上海狮尾智能化科技有限公司
View PDF0 Cites 0 Cited by
Summary
Abstract
Description
Claims
Application Information
AI Technical Summary
This helps you quickly interpret patents by identifying the three key elements:
Problems solved by technology
Method used
Benefits of technology
Problems solved by technology
[0016] Aiming at the problem existing in the prior art that the same features are used for inspection and processing of large targets and small targets, the purpose of the present invention is to provide a pedestrian detection method, system, device and medium based on a deep neural network
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 more
Image
Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
Click on the blue label to locate the original text in one second.
Reading with bidirectional positioning of images and text.
Smart Image
Examples
Experimental program
Comparison scheme
Effect test
Embodiment 1
[0060] A pedestrian detection method based on deep neural network, such as figure 1 As shown, it includes step S1, step S2, step S3, step S4, step S5 and step S6.
[0061] Step S1, obtain the original image training set, the original image training set contains multiple original images, and then divide a plurality of sub-images from the distant imaging of each original image according to the perspective projection relationship to form the sub-image training set;
[0062] Among them, the original images in the original image training set are all taken from the ImageNet database. Such as figure 2 As shown, the sub-picture is taken from the upper half of the height of the original picture (that is, the foreground position of the picture), and when taking the sub-picture, make the aspect ratio of the sub-picture consistent with the original picture, and the multiple sub-pictures have a certain distance between each other. degree of overlap to avoid objects appearing at the edge...
Embodiment 2
[0105] A pedestrian detection system based on deep neural network, such as Figure 6 shown, including
[0106] The acquisition module is used to acquire the original image and the input image;
[0107] The segmentation module is used to segment a plurality of subimages from the distant imaging of each original image according to the perspective projection relationship, and is used to segment a plurality of small target pictures from the distant imaging of the input image according to the perspective projection relationship;
[0108] The scaling module is used to scale the sub-image and the original image to a uniform size, and is used to scale the small target image and the input image to a uniform size;
[0109] The training module is used to input the sub-image and the original image into the Faster R-CNN model for training, and obtain the sub-image pedestrian detection model and the original image pedestrian detection model;
[0110] The detection module is used to perfor...
Embodiment 3
[0113] an electronic device such as Figure 7 shown, including
[0114] memory storing executable program code; and
[0115] a processor coupled to the memory;
[0116] Wherein, the processor invokes the executable program code stored in the memory to execute the steps of the pedestrian detection method based on the deep neural network in the first embodiment.
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 pedestrian detection method and system based on a deep neural network, equipment and a medium, and belongs to the field of computer vision, and the method comprises the steps: S1, obtaining an original image, and segmenting a plurality of sub-images from a remote image of the original image according to a perspective projection relation to form a sub-image training set; s2, zooming the original image and the corresponding sub-images to a uniform size; s3, respectively inputting the sub-image and the original image into a Faster R-CNN model for training, and obtaining a sub-image pedestrian detection model and an original image pedestrian detection model; s4, acquiring an input picture and segmenting a plurality of small target pictures from the distant image of the input picture according to the perspective projection relation; s5, respectively detecting the small target picture and the input picture by using the two detection models to obtain two types of detection results; and S6, combining the two types of detection results by using a non-maximum suppression method to obtain a final detection result. According to the method and system, target features can be learned by using different deep networks for different pictures, and the training model is obtained, so that the detection is more targeted, and the effect is better.
Description
technical field [0001] The present invention relates to the field of computer vision, in particular to a pedestrian detection method, system, device and medium based on a deep neural network. Background technique [0002] Traditional pedestrian detection methods use traditional manual features for pedestrian detection, such as: [0003] Deformable part-based models (Pedro F Felzenszwalb, RossB Girshick, David McAllester, and Deva Ramanan. Object detection with discriminatively trained part-based models. TPAMI, 32(9), pp. 1627-1645, 2010.) , which is to detect each part of the target, so as to achieve the purpose of detecting the target; and [0004] Integral Channel Features (ICF) (Piotr Dollar, Zhuowen Tu, Pietro Perona, and Serge Belongie. Integral channel features. In BMVC, volume2, pp. 5, 2009.) is a very widely used pedestrian detection method, the channel feature pyramid and enhanced training classifier are used here; [0005] Aggregated Channel Features (ACF) (Piot...
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
Application Date:The date an application was filed.
Publication Date:The date a patent or application was officially published.
First Publication Date:The earliest publication date of a patent with the same application number.
Issue Date:Publication date of the patent grant document.
PCT Entry Date:The Entry date of PCT National Phase.
Estimated Expiry Date:The statutory expiry date of a patent right according to the Patent Law, and it is the longest term of protection that the patent right can achieve without the termination of the patent right due to other reasons(Term extension factor has been taken into account ).
Invalid Date:Actual expiry date is based on effective date or publication date of legal transaction data of invalid patent.
Login to view more 
Login to view more