Patent
Rapid object recognition method for mobile robot based on deep learning
TLDR
In this paper, a rapid object recognition method for a mobile robot based on deep learning is proposed, and the method comprises the following steps: 1) obtaining of a mobile image; 2), the preprocessing of image data; 3), extracting of the features of the image; 4), image prediction output; 5), environment constrained optimization; 6), image recognition output.Abstract:
The invention discloses a rapid object recognition method for a mobile robot based on deep learning, and the method comprises the following steps: 1), the obtaining of a mobile image; 2), the preprocessing of image data; 3), the extracting of the features of the image; 4), image prediction output; 5), environment constrained optimization; 6), image recognition output. Through the unified integration of detection and recognition results, the method overcomes the shortcomings of the complexity and instability of a conventional object recognition system which needs to employ an object. Through a multilayer residual error network design and the generation of environment gravity constraint conditions, the method irons out the defect that an integrated object recognition system is poor in accuracy. The integration of detection and recognition tasks can guarantee the processing efficiency of a system, and improves the sensing capability of the robot in a moving process.read more
Citations
More filters
Patent
Cleaning robot and controlling method thereof
TL;DR: In this article, a cleaning robot including a driving unit, an obstacle sensor, and a controller is configured to reduce the driving speed of the cleaning robot when the robot contacts an obstacle.
Patent
Facial image age estimation method based on three-level residual error network
TL;DR: Zhang et al. as mentioned in this paper proposed a facial image age estimation method based on a three-level residual error network, which aims to increase the age estimation level under a non-limited condition.
Patent
Human face skin attribute identification method and apparatus, electronic device and storage medium
TL;DR: Wang et al. as discussed by the authors fused features extracted by at least one relatively shallow convolutional layer in the neural network with features extracted from the final convolution layer to obtain a fused feature and based on the fused feature, performing skin attribute prediction on the human face image to acquire a skin attribute tag.
Patent
Barrier segmentation method and device, computer equipment and readable medium
Xun Sun,Xie Yuanfan,Wang Liang +2 more
TL;DR: In this article, a method for barrier segmentation is proposed, which includes the following steps of obtaining multiple sets of window pairs belonging to one preset neighborhood interval from a plurality of windows acquired according to barrier point cloud around current vehicles; obtaining characteristic information of two windows in each set of windows pairs and predicting semantic characteristic information corresponding to center points of the two windows of a corresponding window pair according to the characteristic information, and a pre-trained semantic characteristic model.
Patent
Depth residual network construction method for mobile phone screen defect detection
TL;DR: In this paper, a depth residual network construction method for mobile phone screen defect detection is proposed, which comprises steps: 1) pictures with defects and normal pictures are acquired respectively, marking is carried out (1 presents a picture with a defect, and 0 presents a normal pictu re), and a user-defined depth residual networks is trained through training data until convergence and high accuracy; 2) a method of using a certain probability for random removal for each residual module in the depth residual net, a shallow network model is generated, the operation is repeated, and multiple network models with
References
More filters
Patent
Internet video face recognition method based on deep learning
TL;DR: In this article, an Internet video face recognition method based on deep learning is proposed, which consists of marking face data, acquiring an image with a face from the Internet, marking a face location box and the name of the person, and establishing a face image library.
Patent
Firework identification method and firework identification system based on deep learning of image
TL;DR: Wang et al. as discussed by the authors proposed a firework identification method based on deep learning of an image, which comprises the following steps of step 1, acquiring a label-free sample image set and a label image set; step 2, performing whitening preliminary processing on training data; step 4, based on the label free training data subjected to the whitened preliminary processing, constructing a deep neutral network based on sparse self coding by adopting unsupervised learning, and extracting a basic image feature set of the labelfree training data, convolving basic image features and pooling image data.
Patent
Video moving target classification and identification method based on outline constraint
TL;DR: In this paper, a video moving target classification and identification method based on outline constraint is proposed, which includes the steps of obtaining a realistic target region and a target outline through a level set partitioning algorithm which is based on color features, textural features and shape prior constraint.
Patent
Real-time target detection method based on oriented gradient two-value mode and soft cascade SVM
TL;DR: In this paper, the authors proposed a real-time target detection method based on the oriented gradient two-value mode (ORBP) and soft cascade SVM, which aims at solving the problems that the target detection in the prior art is low in instantaneity and robustness.
Related Papers (5)
Object detection and recognition of intelligent service robot based on deep learning
Yanan Zhang,Hongyu Wang,Fang Xu +2 more