A Pedestrian Detection Method Based on Fast Construction of Image Feature Pyramid
A technology for image feature and pedestrian detection, which is applied in biometric recognition, character and pattern recognition, instruments, etc., can solve the problems of increased calculation amount, low recognition speed and low calculation amount, and achieve the effect of improving efficiency and reducing calculation amount
Active Publication Date: 2019-02-26
四川御刃科技有限公司
View PDF3 Cites 0 Cited by
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
However, as the accuracy increases, it is often accompanied by an increase in the amount of calculation
However, for most applications in visual recognition (such as robotics, human-computer interaction, automotive safety, and mobile devices), faster recognition speed and lower computational load are always required
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] The present invention will be described in detail below with an example of pedestrian detection. It should be noted that the described embodiment is only intended to facilitate the understanding of the present invention, and does not have any limiting effect on it.
[0030] In order to realize the efficient execution of the pedestrian detection algorithm, in the preferred embodiment of the present invention, for a given pedestrian image, multiple channel features of the image are calculated, and when constructing the image feature pyramid, the image features of the adjacent scale are used to directly perform Approximate estimation, and then detection based on the obtained image features. It is characterized in that, comprising the following steps:
[0031] S1. For a given pedestrian image I, calculate multiple channel features C of the pedestrian image I, where C=Ψ(I), Ψ is the process of calculating the channel features of the pedestrian image. Among them, the channel...
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 based on quickly constructing an image feature pyramid, which includes: S1, calculating multiple channel features C of the pedestrian image; S2, performing fast image features on the multiple channel features C described in step S1 Pyramid construction, and obtain image features of multiple scales; S3, build a strong classifier required; S4, use a fixed-size window to perform sliding scanning on the image features of multiple scales obtained in S2, each sliding At this time, according to the strong classifier obtained in S3, the pedestrian detection is performed on the position of the window at this time; S5, the detection windows are merged, and finally a detection result is obtained. The present invention directly approximates the image features of other scales by using the image features of adjacent scales, so as to significantly reduce the calculation amount of the detector and greatly improve the efficiency of pedestrian detection.
Description
technical field [0001] The invention relates to the field of pattern recognition and digital image processing, in particular to a pedestrian detection method based on fast construction of an image feature pyramid. Background technique [0002] Decomposing an image with multiple resolutions and multiple angles is a basic image analysis method. Obviously, when the method of representing an image is over-complete, it will be easier to extract the required visual information from it. In other words, the accuracy of the information obtained through calculation is more accurate at this time. For example, it is a very effective method to construct an image feature pyramid as an information source for target recognition, and to improve the recognition accuracy through step-by-step calculations. However, as the accuracy increases, it is often accompanied by an increase in the amount of calculation. However, for most applications in visual recognition (such as robotics, human-compu...
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): G06K9/00G06K9/46
CPCG06V40/10G06V10/56
Inventor 邹刚蒋涛
Owner 四川御刃科技有限公司