CN108734123B - Highway sign recognition method, electronic device, storage medium, and system - Google Patents
Highway sign recognition method, electronic device, storage medium, and system Download PDFInfo
- Publication number
- CN108734123B CN108734123B CN201810478622.3A CN201810478622A CN108734123B CN 108734123 B CN108734123 B CN 108734123B CN 201810478622 A CN201810478622 A CN 201810478622A CN 108734123 B CN108734123 B CN 108734123B
- Authority
- CN
- China
- Prior art keywords
- image
- feature
- gradient
- traffic sign
- highway
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/56—Context or environment of the image exterior to a vehicle by using sensors mounted on the vehicle
- G06V20/58—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads
- G06V20/582—Recognition of moving objects or obstacles, e.g. vehicles or pedestrians; Recognition of traffic objects, e.g. traffic signs, traffic lights or roads of traffic signs
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
- G06F18/2148—Generating training patterns; Bootstrap methods, e.g. bagging or boosting characterised by the process organisation or structure, e.g. boosting cascade
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/56—Extraction of image or video features relating to colour
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Evolutionary Computation (AREA)
- Evolutionary Biology (AREA)
- General Engineering & Computer Science (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Image Analysis (AREA)
- Traffic Control Systems (AREA)
Abstract
本发明提供高速公路标志识别方法，包括步骤：获取待识别图像，多尺度缩放图像，生成特征图像，滑动遍历图像，获取高速公路交通标志的RGB图像和深度图像；将深度图像的深度信息映射至RGB图像，生成融合图像，对融合图像进行多尺度缩放，生成若干缩放图像；计算缩放图像的包含颜色信息和梯度信息的特征图像，计算特征图像的积分图像，采用特征选择分类器选择积分图像的特征，生成选择特征图像；通过滑动窗口对选择特征图像进行遍历，识别滑动窗口内的交通标志。本发明还涉及存储介质、电子设备、高速公路标志识别系统，本发明采用自动图像识别的方式对高速公路交通标志自动判别，提高了车辆在高速公路行驶过程中的交通标志识别的工作效率及识别精度。
The present invention provides a highway sign recognition method, which includes the steps of: acquiring an image to be recognized, scaling the image at multiple scales, generating a feature image, sliding through the image, and acquiring an RGB image and a depth image of a highway traffic sign; mapping the depth information of the depth image to RGB image, generate a fused image, perform multi-scale scaling on the fused image, and generate several zoomed images; feature to generate a selected feature image; traverse the selected feature image through a sliding window to identify the traffic signs in the sliding window. The invention also relates to a storage medium, an electronic device, and a highway sign recognition system. The invention adopts an automatic image recognition method to automatically discriminate the highway traffic signs, thereby improving the work efficiency and recognition of the traffic sign recognition during the driving process of the vehicle on the highway. precision.
Description
Technical Field
The present invention relates to the field of traffic sign recognition technologies, and in particular, to a method, an electronic device, a storage medium, and a system for recognizing a highway sign.
Background
With the scientific progress and the development of urbanization, the number of vehicles on the highway and the number of people going out are greatly increased, and the problem of high-speed traffic safety becomes increasingly prominent. The intelligent transportation system has wide economic benefits and profound social influence. Traffic sign recognition systems, as an important subsystem of intelligent vehicles, have become an important component of semi-automatic and automatic vehicles. The real-time requirement of the traffic sign identification of the highway is high, when the traffic sign is detected, the input image is a video shot by a vehicle in the running process of the highway, and the existing highway traffic sign identification has the following problems: videos shot in the driving process of the expressway contain a large amount of environmental information, and the traffic signs cannot be stably and effectively detected in an environment with mixed backgrounds; traffic signs are generally small, various in types and large in sign difference, and detection difficulty is increased; the outdoor scene is often illuminated and shielded, and the detection difficulty is increased; the shot image is large, the resolution is high, and a large amount of time is consumed for processing the image, so that the real-time performance is poor.
Disclosure of Invention
In order to overcome the defects of the prior art, one of the purposes of the invention is to provide a highway sign identification method, which adopts an automatic image identification mode to automatically judge the traffic sign of the highway, and improves the working efficiency and the identification precision of the traffic sign identification of vehicles in the running process of the highway.
The invention provides a highway sign identification method, which comprises the following steps:
acquiring an image to be identified, and acquiring an RGB image and a depth image of a highway traffic sign;
the method comprises the steps of multi-scale zooming images, mapping depth information of the depth images to the RGB images to generate fusion images, and carrying out multi-scale zooming on the fusion images to generate a plurality of zooming images;
generating a characteristic image, calculating the characteristic image containing color information and gradient information of the zoomed image, calculating an integral image of the characteristic image, and generating a selection characteristic image by taking the characteristics of the integral image as the input of a characteristic selection classifier;
and sliding the traversal image, traversing the selected characteristic image through a sliding window, and identifying the traffic sign in the sliding window.
Further, the method also comprises the following steps of extracting a traffic sign area: dividing the selected characteristic image into a plurality of sub-regions along the width direction, traversing each sub-region, calculating the gradient amplitude and the gradient angle of each sub-region, calculating the gradient amplitude sum of the gradient angle in each sub-region from +/-1 DEG to +/-6 DEG, and generating a traffic sign region in the selected characteristic image.
Further, the step of extracting the traffic sign region further comprises: and performing secondary area extraction on the traffic sign area by adopting the depth value, acquiring an area with a fixed value of the depth value change rate in the length direction, and generating an optimized traffic sign area.
Further, the step of generating the feature image further includes estimating gradient integral features of the scaled image of different scales by using an interpolation algorithm, generating the integral image according to the gradient integral features, and selecting the features of the integral image by using a feature selection classifier with an increasing number of layers to generate a selected feature image.
And further, detecting context information, namely inputting the spatial context information and the temporal context information of the fused image into the feature selection classifier, and selecting the traffic sign with the highest feature score in the overlapped boxes.
An electronic device, comprising: a processor;
a memory; and a program, wherein the program is stored in the memory and configured to be executed by the processor, the program comprising instructions for performing the highway sign recognition method described above.
A computer-readable storage medium, on which a computer program is stored, which computer program is executed by a processor to carry out the above-mentioned highway sign recognition method.
A highway sign identification system comprising:
the module for acquiring the image to be identified comprises: the system comprises a display device, a display device and a display device, wherein the display device is used for displaying an image of the traffic sign of the expressway;
multi-scale scaling the image module: the depth information of the depth image is mapped to the RGB image to generate a fusion image, and the fusion image is subjected to multi-scale scaling to generate a plurality of scaling images;
a feature image generation module: the feature image which is used for calculating the zoom image and contains color information and gradient information, the integral image of the feature image is calculated, and the feature of the integral image is used as the input of a feature selection classifier to generate a selection feature image;
sliding and traversing the image module: and traversing the selected characteristic image through a sliding window, and identifying the traffic sign in the sliding window.
Further, the module for extracting the traffic sign area comprises the following modules: the system comprises a selection characteristic image, a gradient amplitude value and a gradient angle, a gradient amplitude value sum of the gradient angle in each sub-region from +/-1 degrees to +/-6 degrees, and a traffic sign region in the selection characteristic image, wherein the selection characteristic image is divided into a plurality of sub-regions along the width direction, each sub-region is traversed, the gradient amplitude value and the gradient angle of each sub-region are calculated, and the traffic sign region in the selection characteristic image is generated; the module for extracting the traffic sign area further comprises the step of performing secondary area extraction on the traffic sign area by adopting the depth value, acquiring an area with a fixed depth value change rate along the length direction, and generating an optimized traffic sign area.
Further, the feature image generation module further estimates gradient integral features of the scaled images of different scales by adopting an interpolation algorithm, generates the integral image according to the gradient integral features, and selects the features of the integral image by adopting a feature selection classifier with increasing layer number to generate a selected feature image; the system also comprises a context information detection module: and the system is used for inputting the spatial context information and the temporal context information of the fused image into the feature selection classifier and selecting the traffic sign with the highest feature score in the overlapped boxes.
Compared with the prior art, the invention has the beneficial effects that:
the invention provides a highway sign identification method, which comprises the following steps: acquiring an image to be identified, scaling the image in multiple scales, generating a characteristic image, traversing the image in a sliding manner, and acquiring an RGB (red, green and blue) image and a depth image of a highway traffic sign; mapping the depth information of the depth image to an RGB image to generate a fusion image, and carrying out multi-scale scaling on the fusion image to generate a plurality of scaled images; calculating a feature image containing color information and gradient information of the zoomed image, calculating an integral image of the feature image, and generating a selection feature image by taking the features of the integral image as the input of a feature selection classifier; and traversing the selected characteristic image through the sliding window, and identifying the traffic sign in the sliding window. The invention also relates to a storage medium, electronic equipment and a highway sign identification system, and the invention adopts an automatic image identification mode to automatically judge the highway traffic sign, thereby improving the working efficiency and the identification precision of the traffic sign identification of vehicles in the running process of the highway.
The foregoing description is only an overview of the technical solutions of the present invention, and in order to make the technical solutions of the present invention more clearly understood and to implement them in accordance with the contents of the description, the following detailed description is given with reference to the preferred embodiments of the present invention and the accompanying drawings. The detailed description of the present invention is given in detail by the following examples and the accompanying drawings.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the invention and together with the description serve to explain the invention without limiting the invention. In the drawings:
FIG. 1 is a flow chart of a highway marking identification method of the present invention;
fig. 2 is a schematic structural diagram of the highway sign recognition system of the present invention.
Detailed Description
The present invention will be further described with reference to the accompanying drawings and the detailed description, and it should be noted that any combination of the embodiments or technical features described below can be used to form a new embodiment without conflict.
The method for identifying the highway sign, as shown in fig. 1, comprises the following steps:
acquiring an image to be identified, and acquiring an RGB image and a depth image of a highway traffic sign; in this embodiment, an RGB image acquired by an industrial camera is acquired, and a depth image acquired by a depth camera is acquired, where the RGB image and the depth image are images acquired by the same target at the same time. When the image is acquired, the industrial camera and the depth camera are in the same acquisition size, the image at the same position is acquired at the same time, in order to keep the images acquired by the two cameras to eliminate the visual angle difference as much as possible, the industrial camera and the depth camera are vertically arranged, and the lens direction is the horizontal direction.
And (3) multi-scale zooming the image, mapping the depth information of the depth image to the RGB image to generate a fusion image, and carrying out multi-scale zooming on the fusion image to generate a plurality of zoomed images.
Generating a characteristic image, calculating the characteristic image containing color information and gradient information of the zoomed image, calculating an integral image of the characteristic image, and generating a selection characteristic image by taking the characteristics of the integral image as the input of a characteristic selection classifier; in one embodiment, the original RGB image is color-converted into a LUV space with a low coupling degree, and 3 color feature images, 6 gradient feature images in different directions, and l gradient images including gradient size are obtained by calculation; calculating an integral image corresponding to each characteristic image; randomly selecting any one feature image, rectangular frames with any position and any size, and quickly calculating an integral value in the rectangular frames through an integrogram to obtain the l-dimensional features. And (3) performing feature selection by adopting an Adaboost algorithm, obtaining a final score through the weighted sum of a plurality of weak classifiers, generating a strong classifier, performing secondary classification, and realizing that detection is completed by only using a small amount of features in actual detection.
In one embodiment, the method further comprises the step of extracting the traffic sign area: dividing the selected characteristic image into a plurality of sub-regions along the width direction, traversing each sub-region, calculating the gradient amplitude and the gradient angle of each sub-region, calculating the gradient amplitude sum of the gradient angle in each sub-region from +/-1 degrees to +/-6 degrees, and generating the traffic sign region in the selected characteristic image. The step of extracting the traffic sign area further comprises: and performing secondary area extraction on the traffic sign area by adopting the depth value, acquiring the area with the depth value change rate in the length direction as a fixed value, and generating an optimized traffic sign area, wherein the depth value is check information, and because the depth value in the background changes into nonlinear change and the depth value in the length direction on the traffic sign continuously changes or keeps unchanged, performing second-order derivation on the depth value, and when the second-order derivation of the depth value change rate is zero, assisting in verifying the correctness of the traffic sign area extracted in the area.
In an embodiment, the information of the gradient does not have scale invariance, and the proportional relation of the gradient integral values in the same area cannot be directly estimated in images with different scaling sizes; the method comprises the following steps of amplifying an original image into an up-sampling image, reducing the original image into a down-sampling image, and estimating integral gradient information of the up-sampling image and the down-sampling image according to the following formula:
wherein, I is an original image; f (I, S) is the original image scaling k is 2SThen, estimating a formula by using the gradient integral characteristic; when S is>0，k>1 denotes an up-sampled image; when S is<0，k<1 denotes a down-sampled image.
In an embodiment, preferably, the step of generating the feature image further includes selecting features of the integral image by using a feature selection classifier with an increasing number of layers to generate the selected feature image. In order to process the traffic sign detection task with a large number of negative samples, a simple classifier is used for roughly filtering the samples to filter out samples which are obviously not marked; and then fine filtering is performed by a complex classifier. Specifically, an Adaboost classifier with the maximum number of layers being T is adopted, and the complexity and the classification capability of the classifier are increased through the Adaboost classifier with the number of layers increasing, so that a large number of negative samples are rejected in the early stage.
And sliding the traversal image, traversing the selected characteristic image through a sliding window, and detecting the traffic sign in the sliding window.
In one embodiment, the spatial context information includes location, size, width; the time context information is the image, the horizontal axis coordinate, the vertical axis coordinate and the width of a sliding window in the original image which needs to be detected by the current frame, preferably, the method also comprises the step of detecting the context information, inputting the space context information and the time context information of the fused image into a feature selection classifier, and selecting the traffic sign with the highest feature score in the overlapped boxes.
An electronic device, comprising: a processor; a memory; and a program, wherein the program is stored in the memory and configured to be executed by the processor, the program comprising instructions for performing the highway sign identification method described above; a computer-readable storage medium, on which a computer program is stored, the computer program being executed by a processor for the above-mentioned highway marking identification method.
The highway sign recognition system, as shown in fig. 2, includes:
the image to be recognized is acquired by the image acquiring module to acquire the RGB image and the depth image of the highway traffic sign.
The multi-scale image scaling module maps the depth information of the depth image to the RGB image to generate a fusion image, and performs multi-scale scaling on the fusion image to generate a plurality of scaling images.
The feature image generation module calculates a feature image containing color information and gradient information of the zoomed image, calculates an integral image of the feature image, and generates a selection feature image by taking the features of the integral image as the input of the feature selection classifier; in one embodiment, the feature image generation module performs color conversion on the original RGB image into an LUV space with a low coupling degree, and calculates to obtain 3 color feature images, 6 gradient feature images in different directions, and l gradient images including gradient sizes; calculating an integral image corresponding to each characteristic image; randomly selecting any one feature image, rectangular frames with any position and any size, and quickly calculating an integral value in the rectangular frames through an integrogram to obtain the l-dimensional features. And the characteristic image generation module adopts an Adaboost algorithm to perform characteristic selection, obtains a final score through the weighted sum of a plurality of weak classifiers, generates a strong classifier, performs secondary classification, and realizes that detection is completed by only using a small amount of characteristics in actual detection.
In one embodiment, the method further comprises the step of extracting the traffic sign area module: dividing the selected characteristic image into a plurality of sub-regions along the width direction, traversing each sub-region, calculating the gradient amplitude and the gradient angle of each sub-region, calculating the gradient amplitude sum of the gradient angle in each sub-region from +/-1 degrees to +/-6 degrees, and generating the traffic sign region in the selected characteristic image. The module for extracting the traffic sign area further comprises: and performing secondary area extraction on the traffic sign area by adopting the depth value, acquiring the area with the depth value change rate in the length direction as a fixed value, and generating an optimized traffic sign area, wherein the depth value is check information, and because the depth value in the background changes into nonlinear change and the depth value in the length direction on the traffic sign continuously changes or keeps unchanged, performing second-order derivation on the depth value, and when the second-order derivation of the depth value change rate is zero, assisting in verifying the correctness of the traffic sign area extracted in the area.
In an embodiment, the information of the gradient does not have scale invariance, and the proportional relation of the gradient integral values in the same area cannot be directly estimated in images with different scaling sizes; the method comprises the following steps of amplifying an original image into an up-sampling image, reducing the original image into a down-sampling image, and estimating integral gradient information of the up-sampling image and the down-sampling image according to the following formula:
wherein, I is an original image; f (I, S) is the original image scaling k is 2SThen, estimating a formula by using the gradient integral characteristic; when S is>0，k>1 denotes an up-sampled image; when S is<0，k<1 denotes a down-sampled image.
In an embodiment, preferably, the feature image generation module selects features of the integral image by using a feature selection classifier with an increasing number of layers to generate the selected feature image. In order to process the traffic sign detection task with a large number of negative samples, a characteristic image module is selected to firstly carry out coarse filtration on the samples by using a simple classifier, and the samples which are obviously not signs are filtered; and then fine filtering is performed by a complex classifier. Specifically, an Adaboost classifier with the maximum number of layers being T is adopted, and the complexity and the classification capability of the classifier are increased through the Adaboost classifier with the number of layers increasing, so that a large number of negative samples are rejected in the early stage.
And the sliding traversal image module traverses the selected characteristic image through the sliding window and detects the traffic sign in the sliding window.
In one embodiment, the spatial context information includes location, size, width; the time context information is the image, the horizontal axis coordinate, the vertical axis coordinate and the width of a sliding window in the original image to be detected of the current frame, preferably, the method further comprises the step that a context information detection module inputs the space context information and the time context information of the fused image into a feature selection classifier, and selects the traffic sign with the highest feature score in the overlapped boxes.
The invention provides a highway sign identification method, which comprises the following steps: acquiring an image to be identified, scaling the image in multiple scales, generating a characteristic image, traversing the image in a sliding manner, and acquiring an RGB (red, green and blue) image and a depth image of a highway traffic sign; mapping the depth information of the depth image to an RGB image to generate a fusion image, and carrying out multi-scale scaling on the fusion image to generate a plurality of scaled images; calculating a feature image containing color information and gradient information of the zoomed image, calculating an integral image of the feature image, and generating a selection feature image by taking the features of the integral image as the input of a feature selection classifier; and traversing the selected characteristic image through the sliding window, and identifying the traffic sign in the sliding window. The invention also relates to a storage medium, electronic equipment and a highway sign identification system, and the invention adopts an automatic image identification mode to automatically judge the highway traffic sign, thereby improving the working efficiency and the identification precision of the traffic sign identification of vehicles in the running process of the highway.
The foregoing is merely a preferred embodiment of the invention and is not intended to limit the invention in any manner; those skilled in the art can readily practice the invention as shown and described in the drawings and detailed description herein; however, those skilled in the art should appreciate that they can readily use the disclosed conception and specific embodiments as a basis for designing or modifying other structures for carrying out the same purposes of the present invention without departing from the scope of the invention as defined by the appended claims; meanwhile, any changes, modifications, and evolutions of the equivalent changes of the above embodiments according to the actual techniques of the present invention are still within the protection scope of the technical solution of the present invention.
Claims (7)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810478622.3A CN108734123B (en) | 2018-05-18 | 2018-05-18 | Highway sign recognition method, electronic device, storage medium, and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810478622.3A CN108734123B (en) | 2018-05-18 | 2018-05-18 | Highway sign recognition method, electronic device, storage medium, and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108734123A CN108734123A (en) | 2018-11-02 |
| CN108734123B true CN108734123B (en) | 2021-09-17 |
Family
ID=63937626
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810478622.3A Expired - Fee Related CN108734123B (en) | 2018-05-18 | 2018-05-18 | Highway sign recognition method, electronic device, storage medium, and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108734123B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110096981A (en) * | 2019-04-22 | 2019-08-06 | 长沙千视通智能科技有限公司 | A kind of video big data traffic scene analysis method based on deep learning |
| CN110097600B (en) * | 2019-05-17 | 2021-08-06 | 百度在线网络技术（北京）有限公司 | Method and device for identifying traffic sign |
| CN111768415A (en) * | 2020-06-15 | 2020-10-13 | 哈尔滨工程大学 | Image instance segmentation method without quantization pooling |
| CN114742912A (en) * | 2020-12-23 | 2022-07-12 | 智道网联科技(北京)有限公司 | Method and device for generating guideboard |
| CN115393828A (en) * | 2022-09-22 | 2022-11-25 | 贵州宽凳智云科技有限公司 | Signboard identification method and device, computer equipment and storage medium |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103390167A (en) * | 2013-07-18 | 2013-11-13 | 奇瑞汽车股份有限公司 | Multi-characteristic layered traffic sign identification method |
| CN105930830A (en) * | 2016-05-18 | 2016-09-07 | 大连理工大学 | A road traffic sign recognition method based on convolutional neural network |
| CN107239730A (en) * | 2017-04-17 | 2017-10-10 | 同济大学 | The quaternary number deep neural network model method of intelligent automobile Traffic Sign Recognition |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8509526B2 (en) * | 2010-04-13 | 2013-08-13 | International Business Machines Corporation | Detection of objects in digital images |
-
2018
- 2018-05-18 CN CN201810478622.3A patent/CN108734123B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103390167A (en) * | 2013-07-18 | 2013-11-13 | 奇瑞汽车股份有限公司 | Multi-characteristic layered traffic sign identification method |
| CN105930830A (en) * | 2016-05-18 | 2016-09-07 | 大连理工大学 | A road traffic sign recognition method based on convolutional neural network |
| CN107239730A (en) * | 2017-04-17 | 2017-10-10 | 同济大学 | The quaternary number deep neural network model method of intelligent automobile Traffic Sign Recognition |
Non-Patent Citations (1)
| Title |
|---|
| 支持无人驾驶车辆的交通标志检测;俞凌;《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》;20160715(第7期);第3.1-3.2节 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108734123A (en) | 2018-11-02 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108734123B (en) | Highway sign recognition method, electronic device, storage medium, and system | |
| CN105261020B (en) | A kind of express lane line detecting method | |
| CN107392139B (en) | Lane line detection method based on Hough transform and terminal equipment | |
| CN108447303B (en) | Peripheral visual field hazard recognition method based on the coupling of human vision and machine vision | |
| EP3605460A1 (en) | Information processing method and apparatus, cloud processing device and computer program product | |
| CN111695514A (en) | Vehicle detection method in foggy days based on deep learning | |
| CN106971185B (en) | License plate positioning method and device based on full convolution network | |
| CN114067186B (en) | Pedestrian detection method and device, electronic equipment and storage medium | |
| CN114359181A (en) | Intelligent traffic target fusion detection method and system based on image and point cloud | |
| CN109034136A (en) | Image processing method, device, picture pick-up device and storage medium | |
| CN110491132A (en) | Vehicle based on video frame picture analyzing, which is disobeyed, stops detection method and device | |
| CN110733416B (en) | A Lane Departure Warning Method Based on Inverse Perspective Transformation | |
| CN111178150A (en) | Lane line detection method, system and storage medium | |
| KR20200119369A (en) | Apparatus and method for detecting object | |
| CN113255555A (en) | Method, system, processing equipment and storage medium for identifying Chinese traffic sign board | |
| CN107918775B (en) | Zebra crossing detection method and system for assisting safe driving of vehicle | |
| CN107689157A (en) | Traffic intersection based on deep learning can passing road planing method | |
| CN109977862B (en) | Recognition method of parking space limiter | |
| CN112991374A (en) | Canny algorithm-based edge enhancement method, device, equipment and storage medium | |
| CN110348273A (en) | Neural network model training method, system and Lane detection method, system | |
| CN107818583A (en) | Cross searching detection method and device | |
| CN108256487A (en) | A kind of driving state detection device and method based on reversed binocular | |
| CN106228560B (en) | A kind of demographic method under complex scene | |
| Kiran et al. | Automatic hump detection and 3D view generation from a single road image | |
| CN202771439U (en) | Traffic sign automatic identification system based on MATLAB |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20210917 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |