Sparse-representation-LBP-and-HOG-integration-based pedestrian detection method
A sparse representation, pedestrian detection technology, applied in character and pattern recognition, instruments, computer components, etc., can solve the problems of insufficient pedestrian description, high dimension, sparseness, etc., to overcome the lack of description ability, strengthen the description ability, and reduce the dimension. Effect
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Publication Date
- 2016-03-30
Smart Images
Figure 1 Figure 2 Figure 3
Abstract
Description
technical field
[0001] The invention belongs to the field of pedestrian detection under pattern recognition, in particular to a pedestrian detection method based on fusion of sparse representation LBP and HOG. Background technique
[0002] Pedestrian detection can be defined as: judging whether the input picture (or video frame) contains pedestrians, and if so, giving location information. Pedestrian Detection System (PDS-Pedestrian Detection System) aims to establish an autonomous, intelligent pedestrian detection and intelligent assisted driving system on a moving car, which has important significance and practical value in improving driving safety and ensuring the safety of pedestrians' lives and property. In the pedestrian detection system, it usually includes three stages: region of interest extraction, feature extraction, and target recognition.
[0003] The simple features usually extracted for pedestrian detection include the aspect ratio of the target, the duty cyc...
Examples
specific Embodiment approach
[0061] In order to further illustrate the technical solution of the present invention, in conjunction with the accompanying drawings, the specific implementation of the present invention is as follows:
[0062] The invention discloses a pedestrian detection method based on the fusion of sparse representation LBP and HOG. The method first uses training samples to train a classifier model, and then uses the classifier model to identify and detect samples. in:
[0063] like figure 1 As shown, the specific steps of using the training samples to train the classifier model are as follows:
[0064] A1: Input training sample group picture I train ;
[0065] A2: Since the extraction process of LBP features is based on grayscale images, it is judged whether the training sample group pictures are grayscale images, if not, convert them into grayscale images;
[0066] A3: Extract the LBP features of the training sample pictures and perform normalization processing; the specific steps a...