CN112926470A - Pedestrian detection method based on artificial intelligence - Google Patents
Pedestrian detection method based on artificial intelligence Download PDFInfo
- Publication number
- CN112926470A CN112926470A CN202110242394.1A CN202110242394A CN112926470A CN 112926470 A CN112926470 A CN 112926470A CN 202110242394 A CN202110242394 A CN 202110242394A CN 112926470 A CN112926470 A CN 112926470A
- Authority
- CN
- China
- Prior art keywords
- pedestrian
- image
- images
- artificial intelligence
- detection method
- 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.)
- Pending
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 39
- 238000013473 artificial intelligence Methods 0.000 title claims abstract description 24
- 238000000034 method Methods 0.000 claims abstract description 9
- 230000003321 amplification Effects 0.000 claims description 21
- 238000003199 nucleic acid amplification method Methods 0.000 claims description 21
- 238000009434 installation Methods 0.000 claims description 12
- 239000002131 composite material Substances 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 6
- 238000003709 image segmentation Methods 0.000 claims description 6
- 230000000903 blocking effect Effects 0.000 claims description 3
- 238000000605 extraction Methods 0.000 claims description 3
- 230000002194 synthesizing effect Effects 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 12
- 238000005516 engineering process Methods 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/20—Movements or behaviour, e.g. gesture recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
Abstract
The invention discloses a pedestrian detection method based on artificial intelligence, which comprises the steps of acquiring a plurality of images of a road section to be detected through camera equipment, storing the collected images in a database, and forming candidate area images; the method comprises the steps of automatically dividing the body of a pedestrian into different semantic regions through key point information of the body of the pedestrian detected by the camera equipment, and generating a synthetic image by utilizing the semantic regions to input the synthetic image into a pedestrian classification network to finish training of the pedestrian classification network. The invention provides a pedestrian detection method based on artificial intelligence, which avoids the condition that the application scene of the pedestrian detection method based on artificial intelligence is limited, and achieves better use effect in some special scenes, for example, for pedestrian detection in automatic driving, the pedestrian can have the use effect of detection due to the fact that the pedestrian can lie down, or the pedestrian can have the better use effect of detection due to the fact that the pedestrian is blocked by holding an umbrella in the hands in rainy days.
Description
Technical Field
The invention relates to a pedestrian detection method based on artificial intelligence, and belongs to the technical field of artificial intelligence.
Background
Artificial Intelligence (Artificial Intelligence), abbreviated in english as AI. The method is a new technical science for researching and developing theories, methods, technologies and application systems for simulating, extending and expanding human intelligence.
In the prior art, application scenes for the pedestrian detection method based on artificial intelligence are limited, and a good use effect cannot be achieved in some special scenes, for example, for pedestrian detection in automatic driving, the pedestrian cannot achieve the use effect of detection due to the fact that the pedestrian in automatic driving may lie down, or the pedestrian cannot achieve the good use effect of detection due to the fact that the pedestrian blocks a handheld umbrella in rainy days. Therefore, there is a need for a pedestrian detection method based on artificial intelligence to solve the problem in the prior art.
In order to solve the technical problems, a new technical scheme is especially provided.
Disclosure of Invention
The invention aims to provide a pedestrian detection method based on artificial intelligence to solve the problems in the background technology.
In order to achieve the purpose, the invention provides the following technical scheme: an artificial intelligence based pedestrian detection method, the method comprising the steps of:
acquiring a plurality of images of a road section to be detected through camera equipment, and storing the collected images in a database to form a candidate area image;
step two, automatically dividing the body of the pedestrian into different semantic regions through key point information of the body of the pedestrian detected by the camera equipment, generating a synthetic image by using the semantic regions, and inputting the synthetic image into a pedestrian classification network to finish training of the pedestrian classification network; wherein, the key point information of the pedestrian body comprises a body area image;
step three, synthesizing the body area image and the candidate area image into an image at an image level so as to obtain a synthesized image; inputting the synthetic image into a pedestrian classification network to guide the training of the pedestrian classification network;
step four, the pedestrian classification network judges whether the input image is a pedestrian image or a background image according to the input composite image, wherein the composite image comprises semantic area images of all parts of the body of the pedestrian, and the detection of blocking the pedestrian in the real scene is further completed;
step five, amplifying the specified number of pedestrian sample image data containing the specified lying mode in the database in the step one to obtain a first amplification database;
step six, performing pedestrian sample image amplification based on color replacement on the non-lying mode pedestrian sample image data in the first amplification database in the step five to obtain a second amplification database;
and seventhly, carrying out pedestrian sample image amplification based on human body proportion on the non-lying mode pedestrian sample image data in the second amplification database in the sixth step to obtain a pedestrian detection training database.
Preferably, the method for acquiring the candidate region image in the first step includes the following steps:
step 1, extracting the position information of pedestrians and vehicles according to the image acquired by the camera equipment, and removing the pedestrians and vehicles on the selected image;
step 2: extracting image contents of position information corresponding to other images in an image library through computer calculation according to the position information of the pedestrians and the vehicles extracted in the step 1, wherein the image contents at the corresponding positions cannot be the pedestrians and the vehicles, and if not, continuing the extraction until the image information of the pedestrians and the vehicles at the corresponding positions is extracted;
and step 3: filling the image information extracted in the step 2 to the positions of the pedestrians and vehicles removed from the image in the step 1 through computer calculation;
and 4, step 4: and (4) repeating the operation of the step (1-3) until no pedestrian, vehicle or the like exists in the image, and deriving the image to obtain a candidate area image.
Preferably, image segmentation is carried out on all the non-lying mode pedestrian sample images in the first amplification database to obtain an upper loading area and a lower loading area in the non-lying mode pedestrian sample images;
preferably, designated color replacement is carried out on an upper installation area and a lower installation area in the non-lying mode pedestrian sample image after image segmentation, and multiple extended non-lying mode pedestrian sample images with different color combinations of the upper installation area and the lower installation area are obtained.
Preferably, the body area image is a foot and/or leg image.
Preferably, the pedestrian classification network is configured to determine whether the composite image is a pedestrian image or a background image.
Compared with the prior art, the invention has the beneficial effects that: the pedestrian detection method based on the artificial intelligence is provided, the situation that the application scene of the pedestrian detection method based on the artificial intelligence is limited is avoided, a better use effect is achieved in some special scenes, for example, for pedestrian detection in automatic driving, the pedestrian can be in a lying mode to achieve the use effect of detection, or in rainy days, the pedestrian can also achieve the better detection use effect due to the fact that the pedestrian is blocked by a handheld umbrella.
Detailed Description
The technical solutions in the embodiments of the present invention are clearly and completely described below, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The invention provides a technical scheme that: an artificial intelligence based pedestrian detection method, the method comprising the steps of:
acquiring a plurality of images of a road section to be detected through camera equipment, and storing the collected images in a database to form a candidate area image;
step two, automatically dividing the body of the pedestrian into different semantic regions through key point information of the body of the pedestrian detected by the camera equipment, generating a synthetic image by using the semantic regions, and inputting the synthetic image into a pedestrian classification network to finish training of the pedestrian classification network; wherein, the key point information of the pedestrian body comprises a body area image;
step three, synthesizing the body area image and the candidate area image into an image at an image level so as to obtain a synthesized image; inputting the synthetic image into a pedestrian classification network to guide the training of the pedestrian classification network;
step four, the pedestrian classification network judges whether the input image is a pedestrian image or a background image according to the input composite image, wherein the composite image comprises semantic area images of all parts of the body of the pedestrian, and the detection of blocking the pedestrian in the real scene is further completed;
step five, amplifying the specified number of pedestrian sample image data containing the specified lying mode in the database in the step one to obtain a first amplification database;
step six, performing pedestrian sample image amplification based on color replacement on the non-lying mode pedestrian sample image data in the first amplification database in the step five to obtain a second amplification database;
and seventhly, carrying out pedestrian sample image amplification based on human body proportion on the non-lying mode pedestrian sample image data in the second amplification database in the sixth step to obtain a pedestrian detection training database.
Preferably, the method for acquiring the candidate region image in the first step includes the following steps:
step 1, extracting the position information of pedestrians and vehicles according to the image acquired by the camera equipment, and removing the pedestrians and vehicles on the selected image;
step 2: extracting image contents of position information corresponding to other images in an image library through computer calculation according to the position information of the pedestrians and the vehicles extracted in the step 1, wherein the image contents at the corresponding positions cannot be the pedestrians and the vehicles, and if not, continuing the extraction until the image information of the pedestrians and the vehicles at the corresponding positions is extracted;
and step 3: filling the image information extracted in the step 2 to the positions of the pedestrians and vehicles removed from the image in the step 1 through computer calculation;
and 4, step 4: and (4) repeating the operation of the step (1-3) until no pedestrian, vehicle or the like exists in the image, and deriving the image to obtain a candidate area image.
Preferably, image segmentation is carried out on all the non-lying mode pedestrian sample images in the first amplification database to obtain an upper loading area and a lower loading area in the non-lying mode pedestrian sample images;
preferably, designated color replacement is carried out on an upper installation area and a lower installation area in the non-lying mode pedestrian sample image after image segmentation, and multiple extended non-lying mode pedestrian sample images with different color combinations of the upper installation area and the lower installation area are obtained.
Preferably, the body area image is a foot and/or leg image.
Preferably, the pedestrian classification network is configured to determine whether the composite image is a pedestrian image or a background image.
When the pedestrian detection method based on the artificial intelligence is used, the situation that the application scenes of the pedestrian detection method based on the artificial intelligence are limited is avoided, a good use effect is achieved in some special scenes, for example, for pedestrian detection in automatic driving, the pedestrian can be in a lying mode to achieve the use effect of detection, or in rainy days, the pedestrian can be blocked by a handheld umbrella to achieve the good detection use effect.
Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that changes, modifications, substitutions and alterations can be made in these embodiments without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (6)
1. An artificial intelligence based pedestrian detection method, characterized in that the method comprises the following steps:
acquiring a plurality of images of a road section to be detected through camera equipment, and storing the collected images in a database to form a candidate area image;
step two, automatically dividing the body of the pedestrian into different semantic regions through key point information of the body of the pedestrian detected by the camera equipment, generating a synthetic image by using the semantic regions, and inputting the synthetic image into a pedestrian classification network to finish training of the pedestrian classification network; wherein, the key point information of the pedestrian body comprises a body area image;
step three, synthesizing the body area image and the candidate area image into an image at an image level so as to obtain a synthesized image; inputting the synthetic image into a pedestrian classification network to guide the training of the pedestrian classification network;
step four, the pedestrian classification network judges whether the input image is a pedestrian image or a background image according to the input composite image, wherein the composite image comprises semantic area images of all parts of the body of the pedestrian, and the detection of blocking the pedestrian in the real scene is further completed;
step five, amplifying the specified number of pedestrian sample image data containing the specified lying mode in the database in the step one to obtain a first amplification database;
step six, performing pedestrian sample image amplification based on color replacement on the non-lying mode pedestrian sample image data in the first amplification database in the step five to obtain a second amplification database;
and seventhly, carrying out pedestrian sample image amplification based on human body proportion on the non-lying mode pedestrian sample image data in the second amplification database in the sixth step to obtain a pedestrian detection training database.
2. The artificial intelligence based pedestrian detection method of claim 1, wherein: the method for acquiring the candidate area image in the first step comprises the following steps:
step 1, extracting the position information of pedestrians and vehicles according to the image acquired by the camera equipment, and removing the pedestrians and vehicles on the selected image;
step 2: extracting image contents of position information corresponding to other images in an image library through computer calculation according to the position information of the pedestrians and the vehicles extracted in the step 1, wherein the image contents at the corresponding positions cannot be the pedestrians and the vehicles, and if not, continuing the extraction until the image information of the pedestrians and the vehicles at the corresponding positions is extracted;
and step 3: filling the image information extracted in the step 2 to the positions of the pedestrians and vehicles removed from the image in the step 1 through computer calculation;
and 4, step 4: and (4) repeating the operation of the step (1-3) until no pedestrian, vehicle or the like exists in the image, and deriving the image to obtain a candidate area image.
3. The artificial intelligence based pedestrian detection method of claim 1, wherein: and performing image segmentation on all the pedestrian sample images in the non-lying mode in the first amplification database to obtain an upper loading area and a lower loading area in the pedestrian sample images in the non-lying mode.
4. The artificial intelligence based pedestrian detection method of claim 3, wherein: and carrying out appointed color replacement on an upper installation area and a lower installation area in the pedestrian sample image in the non-lying mode after image segmentation to obtain multiple extended non-lying mode pedestrian sample images with different color combinations of the upper installation area and the lower installation area.
5. The artificial intelligence based pedestrian detection method of claim 1, wherein: the body area images are foot and/or leg images.
6. The artificial intelligence based pedestrian detection method of claim 1, wherein: the pedestrian classification network is used for judging whether the composite image is a pedestrian image or a background image.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110242394.1A CN112926470A (en) | 2021-03-05 | 2021-03-05 | Pedestrian detection method based on artificial intelligence |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110242394.1A CN112926470A (en) | 2021-03-05 | 2021-03-05 | Pedestrian detection method based on artificial intelligence |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112926470A true CN112926470A (en) | 2021-06-08 |
Family
ID=76173377
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110242394.1A Pending CN112926470A (en) | 2021-03-05 | 2021-03-05 | Pedestrian detection method based on artificial intelligence |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112926470A (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109766868A (en) * | 2019-01-23 | 2019-05-17 | 哈尔滨工业大学 | A kind of real scene based on body critical point detection blocks pedestrian detection network and its detection method |
| CN110084115A (en) * | 2019-03-22 | 2019-08-02 | 江苏现代工程检测有限公司 | Pavement detection method based on multidimensional information probabilistic model |
| CN110084118A (en) * | 2019-03-25 | 2019-08-02 | 哈尔滨工业大学（深圳） | Method for building up, pedestrian detection method and the device of pedestrian detection tranining database |
| CN110414413A (en) * | 2019-07-25 | 2019-11-05 | 北京麒麟智能科技有限公司 | A kind of logistics trolley pedestrian detection method based on artificial intelligence |
-
2021
- 2021-03-05 CN CN202110242394.1A patent/CN112926470A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109766868A (en) * | 2019-01-23 | 2019-05-17 | 哈尔滨工业大学 | A kind of real scene based on body critical point detection blocks pedestrian detection network and its detection method |
| CN110084115A (en) * | 2019-03-22 | 2019-08-02 | 江苏现代工程检测有限公司 | Pavement detection method based on multidimensional information probabilistic model |
| CN110084118A (en) * | 2019-03-25 | 2019-08-02 | 哈尔滨工业大学（深圳） | Method for building up, pedestrian detection method and the device of pedestrian detection tranining database |
| CN110414413A (en) * | 2019-07-25 | 2019-11-05 | 北京麒麟智能科技有限公司 | A kind of logistics trolley pedestrian detection method based on artificial intelligence |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110335277A (en) | Image processing method, device, computer readable storage medium and computer equipment | |
| CN103810505B (en) | Vehicles identifications method and system based on multiple layer description | |
| Abdollahi et al. | Improving road semantic segmentation using generative adversarial network | |
| CN111145209B (en) | Medical image segmentation method, device, equipment and storage medium | |
| CN109583345B (en) | Road recognition method, device, computer device and computer readable storage medium | |
| CN109801292A (en) | A kind of bituminous highway crack image partition method based on generation confrontation network | |
| CN106709486A (en) | Automatic license plate identification method based on deep convolutional neural network | |
| CN111104903A (en) | Depth perception traffic scene multi-target detection method and system | |
| Maglad | A vehicle license plate detection and recognition system | |
| CN110163109B (en) | Lane line marking method and device | |
| CN108921850B (en) | Image local feature extraction method based on image segmentation technology | |
| CN110070059B (en) | Unstructured road detection method based on domain migration | |
| CN110598560A (en) | Night monitoring and identifying method and system based on neural network enhancement | |
| CN103413323B (en) | Based on the object tracking methods of component-level apparent model | |
| CN105684046A (en) | Generating image compositions | |
| CN113505670A (en) | Remote sensing image weak supervision building extraction method based on multi-scale CAM and super-pixels | |
| Huang et al. | Towards unsupervised single image dehazing with deep learning | |
| Ye et al. | CT-UNet: An improved neural network based on U-Net for building segmentation in remote sensing images | |
| CN100423021C (en) | Method and device for segmentation low depth image | |
| Ng et al. | Scalable Feature Extraction with Aerial and Satellite Imagery. | |
| CN112926470A (en) | Pedestrian detection method based on artificial intelligence | |
| Lee et al. | Enhancement for automatic extraction of RoIs for bone age assessment based on deep neural networks | |
| Dogrusoz et al. | Modeling urban structures using graph-based spatial patterns | |
| CN114494893B (en) | Remote sensing image feature extraction method based on semantic reuse context feature pyramid | |
| CN110428402A (en) | Distorted image recognition methods, device, computer equipment and storage medium |
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 | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20210608 |
|
| RJ01 | Rejection of invention patent application after publication |