Patent
Pedestrian detection method and system based on block occlusion perception
Lei Zhen,Zhang Shifeng,Zhuang Chubin +2 more
- 27 Nov 2018
7
TL;DR: In this article, a pedestrian detection method based on block occlusion perception is proposed, which aims at solving the technical problem of low pedestrian detection accuracy caused by pedestrian occlusions.
read more
Abstract: The invention belongs to the technical field of pattern recognition, in particular to a pedestrian detection method and system based on block occlusion perception, which aims at solving the technicalproblem of low pedestrian detection accuracy caused by pedestrian occlusion. For this purpose, the pedestrian detection method of the invention comprises the following steps: obtaining image featurescorresponding to each preset human body detection area according to a pedestrian image to be detected on the basis of a pre-established pedestrian detection model; performing feature fusion on the obtained image features to obtain the overall feature of the corresponding pedestrian; acquiring a plurality of detection result boxes of the pedestrian image to be detected according to the overall feature; and selecting a detection result box satisfying a preset screening condition from the plurality of obtained detection result boxes. On the basis of the above steps, the pedestrian occluded in theimage to be detected can be effectively detected. The pedestrian detection system in the invention can execute and implement the above method.
read more
Chat with Paper
AI Agents for this Paper
Find similar papers on Google Scholar, PubMed and Arxiv
Write a critical review of this paper
Analyze citations of this paper to find unaddressed research gaps
Citations
Patent
Image recognition method and device
Gao Qian,Cui Xing,Zhang Rongguo,Wang Shaokang,Chen Kuan +4 more
- 13 Mar 2020
TL;DR: In this article, a self-adaptive preprocessing method was proposed to improve the contrast of the image, and enable a lesion region of a suspected lump to be more prominent.
2
Patent
Behavior detection method, system and device and machine readable medium
Zhou Xi,Yao Zhiqiang,Li Jiwei,Jialin Gao,Zhixiang Shi +4 more
- 14 Feb 2020
TL;DR: In this article, a behavior detection method consisting of the steps: obtaining one or more proposal containing targetbehaviors, classifying the proposal containing the target behavior according to one or multiple classifiers to obtain at least two classification results; and combining the at least one classification results to obtain a target behavior detection result.
1
Patent
Shielded target detection method, system and device and storage medium
Yin Fan,Song Wei,Tie Jun,Li Zimao,Mo Haifang,Zhang Huili +5 more
- 10 Sep 2019
TL;DR: In this article, a shielded target detection method, system and device and a storage medium are described, which comprises the steps of judging whether a training set picture has a preset mark or not; when the preset mark exists in the training set image, performing data identification on the image to obtain a shielded level of a target object in the image; adjusting hyper-parameters of a preset detection model according to the shielded level; substituting the training sets pictures into the pre-trained detection model for training to obtain the first detection model; according to to the to-be-det
1
Patent
Image processing method, device and system
Yang Tong
- 05 Apr 2019
TL;DR: In this article, a target image and multiple groups of anchor point information are obtained, where the anchor points information comprises anchor point parameters and category numbers corresponding to anchor points, and the feature map and the classification layer network parameters of the classification prediction network are input into a category prediction network, and determining the category of the target object in the target image.
Patent
Behavior detection method and system, machine readable medium and equipment
Zhou Xi,Yao Zhiqiang,Li Jiwei,Jialin Gao,Zhixiang Shi +4 more
- 14 Feb 2020
TL;DR: In this paper, a machine readable medium and equipment for behavior detection is presented. And the behavior detection method comprises the following steps: acquiring a to-be-detected continuous frame image containing a target behavior; representing the continuous frame images to be detected by using a neural network, so that each frame image can sense the target behavior characteristic information of the current frame image and the target behaviour characteristic information from other frame images; classifying the represented continuous frames images to detect according to a classifier to obtain at least two classification results.
References
•Posted Content
Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
TL;DR: Faster R-CNN as discussed by the authors proposes a Region Proposal Network (RPN) to generate high-quality region proposals, which are used by Fast R-NN for detection.
25.3K
Patent
Object detection and classification in images
Jian Sun,Ross Girshick,Shaoqing Ren,Kaiming He +3 more
- 20 Jan 2016
TL;DR: In this article, a computing device can receive an input image and generate a convolutional feature map, which can then be processed through a Region Proposal Network (RPN) to generate proposals for candidate objects in the image.
110
Patent
Real-time target detection method based on region convolutional neural network
Xia Chunqiu
- 26 Apr 2017
TL;DR: In this article, the authors proposed a real-time target detection method based on a region convolutional neural network (R-CNN) which consists of the steps of: regarding an image of any size as input, inputting a plurality of regions of interest (RoIs) while inputting the image, proposing a detection region by means of a region proposal network(RPN), utilizing the proposed detection regions by an R-CNN detector, learning shared characteristics by adopting alternative training optimization, and carrying out classification detection by using the classifier.
38
Patent
Pedestrian detection method based on end-to-end convolutional neural network
Li Hongsheng,Fan Junming,Zhou Hui,Hu Huan,Cao Bin +4 more
- 12 Oct 2016
TL;DR: In this article, an end-to-end convolutional neural network (CNN) was used for pedestrian detection, and the pedestrian scale and proportion problems were solved by constructing a candidate box regression network, the pyramid technology was not needed, and a lot of computing resources were saved.
34
Patent
Multi-task layered image retrieval method based on depth self-coding convolution neural network
He Xia,Tang Yiping,Wang Liran,Chen Peng,Yuan Gongping,Jin Yujie +5 more
- 09 Feb 2018
TL;DR: In this paper, a multi-task layered image retrieval method based on a depth self-coding convolution neural network is proposed, which is characterized by mainly comprising of a CNN end-to-end for deep learning and training recognition, a rapid visual segmentation detection and positioning method of a region-of-interest secondary screening module based on an RPN network, a coarse search of a full-graph sparse hash code, an area sensing semantic feature and matrix h accurate comparison and search based on the maximum response, and a region of interest selectivitycomparison algorithm
32
Related Papers (5)
Zhou Weisi,Xu Jing,Zhao Rui +2 more
- 20 Oct 2017
Lin Mengxiao,Zhang Xiangyu +1 more
- 23 Nov 2018
Su Xiangbo,Zhong Donghong,Yuan Yuchen +2 more
- 08 Nov 2019
Long Gang,Lin Songwei,Li Fei,Zheng Guizhen +3 more
- 20 Jun 2017