CN111860426A - Single-sample cockpit image signal target detection method and system based on template matching - Google Patents
Single-sample cockpit image signal target detection method and system based on template matching Download PDFInfo
- Publication number
- CN111860426A CN111860426A CN202010752199.9A CN202010752199A CN111860426A CN 111860426 A CN111860426 A CN 111860426A CN 202010752199 A CN202010752199 A CN 202010752199A CN 111860426 A CN111860426 A CN 111860426A
- Authority
- CN
- China
- Prior art keywords
- image
- edge
- template
- cockpit
- cabin
- 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 38
- 238000003708 edge detection Methods 0.000 claims abstract description 72
- 238000012545 processing Methods 0.000 claims abstract description 44
- 238000000034 method Methods 0.000 claims description 26
- 238000007781 pre-processing Methods 0.000 claims description 23
- 238000004364 calculation method Methods 0.000 claims description 14
- 238000005516 engineering process Methods 0.000 claims description 14
- 238000013135 deep learning Methods 0.000 description 13
- 238000012549 training Methods 0.000 description 8
- 238000005286 illumination Methods 0.000 description 4
- 238000003062 neural network model Methods 0.000 description 4
- 230000000694 effects Effects 0.000 description 3
- 238000002372 labelling Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000009432 framing Methods 0.000 description 1
- 238000003709 image segmentation Methods 0.000 description 1
- 230000007257 malfunction Effects 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000005070 sampling Methods 0.000 description 1
- 230000001629 suppression Effects 0.000 description 1
- 230000001360 synchronised effect Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
- 230000003313 weakening effect Effects 0.000 description 1
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
- G06V20/59—Context or environment of the image inside of a vehicle, e.g. relating to seat occupancy, driver state or inner lighting conditions
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- 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
- G06V10/44—Local feature extraction by analysis of parts of the pattern, e.g. by detecting edges, contours, loops, corners, strokes or intersections; Connectivity analysis, e.g. of connected components
-
- 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
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/75—Organisation of the matching processes, e.g. simultaneous or sequential comparisons of image or video features; Coarse-fine approaches, e.g. multi-scale approaches; using context analysis; Selection of dictionaries
- G06V10/751—Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching
-
- 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
- G06V20/59—Context or environment of the image inside of a vehicle, e.g. relating to seat occupancy, driver state or inner lighting conditions
- G06V20/593—Recognising seat occupancy
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computing Systems (AREA)
- Databases & Information Systems (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Image Analysis (AREA)
Abstract
The invention relates to the technical field of image processing, and particularly discloses a single-sample cockpit image signal target detection method based on template matching, which comprises the following steps of: an edge detection step, which is to perform gray level processing on the stored template image and the cabin image acquired in real time, and then perform edge detection on the template image and the cabin image to respectively obtain a template edge image and a cabin edge image; a template matching step, namely detecting a matched part in the cockpit edge image as a target area; and a state judgment step, namely performing color restoration on the target area, calculating the color mean value of the target area after color restoration, and judging the sign state according to the color mean value, wherein the color mean value is the color mean value of a color three channel. A single-sample cockpit image signal target detection system based on template matching is also provided. By adopting the technical scheme of the invention, the running speed of target detection on the cockpit icon can be increased.
Description
Technical Field
The invention relates to the technical field of image processing, in particular to a method and a system for detecting an image signal target of a single-sample cockpit based on template matching.
Background
Object detection, also called object extraction, is an image segmentation technique based on object geometry and statistical features. It processes the detection of the position and the affiliated type of a certain semantic object (such as people, buildings or cars) in the picture in the digital image and video, such as the detection of road signs on roads, the detection of human faces in photos, and the like. Each "target" object that needs to be detected has its own unique features that help detect the location and classification of the target.
Target detection in order to improve the accuracy of recognition, a deep learning method is generally used. The deep learning method acquires massive data and then performs model training on a computer with strong computer capability, so that the obtained deep neural network model has a large amount of data bases, and has good target detection performance and robustness. Therefore, the current target detection technology mostly depends on a deep learning method. For example, chinese patent publication No. CN103528617A discloses an automatic cabin instrument identification and detection method, which includes the following steps: reading in an instrument image; sampling the image; carrying out noise reduction processing on the image by adopting nonlinear vector median filtering; combining a global threshold value method with a local threshold value method, and binarizing the instrument image to obtain a binarized image; thinning the image, accurately detecting a pointer, and forming the thinned pointer into a single-pixel-width pointer; extracting the edge of an instrument by using a medical improved cross vision model; according to the priori knowledge, learning and training are carried out, similar features are searched, and classification comparison is carried out on the instruments; the angle of the pointer is calculated using a gradient method. According to the technical scheme, learning training is carried out according to priori knowledge, similar features are searched, instruments are classified in a contrast mode, and the measurement precision is improved.
However, the deep learning method needs to collect massive data and then perform model training on a computer with powerful computing power, even a supercomputer. For example, an artificial intelligence model is obtained that can be used to identify cockpit signals (e.g., unbelted belts shown on the instrument panel, engine malfunction, excessive water temperature, etc.). It is generally necessary to:
(1) collecting a great amount of cabin images of automobiles with different models and different environments;
(2) for each cockpit image, the specific position and the category of each type of road sign are manually marked;
(3) and designing a deep neural network model, and training the marked car cabin images on a computer with strong computing power for a long time to obtain the deep neural network model.
Therefore, the deep learning mode has the process of labeling and training mass data, and the cost is high.
In addition, the model file obtained by the deep learning method is huge, and is deployed on edge computing equipment, and the edge computing equipment with a GPGPU (general purpose image processor) is required to perform model reasoning. When a new cockpit image signal, such as a lane offset mark of a new automobile, is added to the obtained deep neural network model, the above steps need to be repeated, the model is trained repeatedly, so that the running speed is slow, and the delay of obtaining the target detection result is fatal to the intelligent driving of the automobile.
Disclosure of Invention
In order to solve the technical problem that the running speed of the conventional cabin icon target detection is low, the invention provides a single-sample cabin image signal target detection method and system based on template matching.
The basic scheme of the invention is as follows:
the single-sample cockpit image signal target detection method based on template matching comprises the following steps of:
an edge detection step, which is to perform gray level processing on the stored template image and the cabin image acquired in real time, and then perform edge detection on the template image and the cabin image to respectively obtain a template edge image and a cabin edge image;
a template matching step, namely detecting a matched part in the cockpit edge image as a target area;
and a state judgment step, namely performing color restoration on the target area, calculating the color mean value of the target area after color restoration, and judging the sign state according to the color mean value, wherein the color mean value is the color mean value of a color three channel.
The technical scheme has the advantages that: compared with the traditional deep learning mode, the method and the device have the advantages that the edge detection subarea (the area near the signal) is searched in the template image selected by the frame, so that a huge parameter model required by the deep learning mode is saved, the acquired cockpit image does not need to be trained in the huge parameter model, the high cost caused by the process of marking and training mass data is reduced, and the running speed of target detection on the cockpit icon is increased.
Further, the edge detection step includes an image preprocessing sub-step and an edge detection sub-step,
the image preprocessing substep is used for carrying out image standardization processing, gray level processing and noise reduction processing on the stored template image and the cabin image acquired in real time;
and an edge detection substep, which is used for carrying out edge detection on the template image and the cockpit image which are subjected to image standardization processing and noise reduction processing to respectively obtain a template edge image and a cockpit edge image.
Has the advantages that: the template image and the cockpit image are subjected to image standardization processing and noise reduction processing, the influence of noise on the edge detection precision can be reduced, and the robustness is better.
Further, in the sub-step of edge detection, Canny edge detection technology is adopted to carry out edge detection on the cabin image and the template image.
Has the advantages that: compared with other edge detection technologies, the Canny edge detection technology detects a strong edge and a weak edge respectively by using two different thresholds, and only includes the weak edge in an output image when the weak edge is connected with the strong edge, so that a real weak edge can be detected, and the detection precision is high. And the Canny edge detection technology binarizes the cabin image and the template image, and refines the image, so that the detected edge in the output image is a single-pixel edge, namely, the detection precision is higher, and the robustness is better.
In addition, the Canny edge detection technique does not require a large amount of data to train, which increases the computational rate.
Further, in the template matching step, a CPU is adopted to calculate matched parts of the plurality of template edge images in the cockpit edge image in parallel as target areas.
Has the advantages that: through CPU parallel computation, the method can realize synchronous detection of the parts matched with the template edge images in the cockpit edge images by taking a plurality of template edge images as matching targets, thereby improving the computation speed and ensuring the timeliness of target detection.
Further, in the image preprocessing substep, gaussian blur is used for noise reduction.
Has the advantages that: the Gaussian blur has a good effect of removing the color noise of the image, so that the influence of the image noise and illumination on the cockpit image is greatly reduced, and the robustness is better.
Single sample passenger cabin image signal target detection system based on template matching includes edge computing device, and edge computing device includes edge detection module, template matching module and state judgement module, wherein:
the edge detection module is used for carrying out gray processing on the stored template image and the cabin image acquired in real time, and then carrying out edge detection on the template image and the cabin image to respectively obtain a template edge image and a cabin edge image;
the template matching module is used for capturing a part matched with the template edge image in the cockpit edge image as a target area;
the state judgment module is used for carrying out color restoration on the target area, calculating the color mean value of the target area after color restoration, and then judging the sign state according to the color mean value, wherein the color mean value is the color mean value of the color three channels.
Has the advantages that: 1. according to the scheme, a huge parameter model required by a deep learning mode is omitted, mass data do not need to be collected, the collected mass data do not need to be labeled and trained for a long time, and the images of the passenger cabin only need to be collected once, so that the data labeling and training cost is reduced.
2. Because the files of a huge number of parameter models required by the deep learning mode are huge, the deep learning mode is not suitable for deployment on partial edge computing equipment, and edge computing equipment with a general image processor is required for model reasoning. According to the scheme, a huge number of parameter models required by a deep learning mode are omitted, so that the method is suitable for more edge computing devices, and the deployment of the devices in the edge computing devices is reduced.
3. The method and the device save a huge parameter model required by a deep learning mode, do not need to train the acquired cockpit image in the huge parameter model, and improve the slow running speed of target detection on the cockpit image.
Further, the edge detection module comprises a preprocessing submodule and an edge detection submodule,
the preprocessing submodule is used for carrying out image standardization processing, gray level processing and noise reduction processing on the stored template image and the cabin image acquired in real time;
the edge detection submodule is used for carrying out edge detection on the template image and the cockpit image which are processed by the preprocessing submodule to respectively obtain a template edge image and a cockpit edge image.
Has the advantages that: and the influence of noise on the edge detection precision is reduced.
Further, the edge detection submodule performs edge detection on the stored template image and the cabin image acquired in real time by adopting a Canny edge detection technology.
Has the advantages that: the Canny edge detection technology uses two different thresholds to detect a strong edge and a weak edge respectively, and when the weak edge and the strong edge are connected, the weak edge is contained in an output image, so that a true weak edge can be detected, and the detection precision is high. And the Canny edge detection technology binarizes the cabin image and the template image, and refines the image, so that the detected edge in the output image is a single-pixel edge, namely, the detection precision is higher, and the robustness is better.
In addition, the Canny edge detection technique does not require a large amount of data to train, which increases the computational rate.
Further, the template matching module captures a portion matching the plurality of template edge images in the cockpit edge image as a target area by using a CPU for parallel computation.
Has the advantages that: the matching calculation tasks of different template edge images are matched through parallel calculation, the calculation speed is improved, and the timeliness of target detection is guaranteed.
Further, the preprocessing submodule performs noise reduction processing on the received template image and the cabin image by adopting Gaussian blur.
Has the advantages that: gaussian blur has a good effect of removing image color noise, so that the influence of image noise and illumination on a cockpit image is greatly reduced, and the robustness is better.
Drawings
FIG. 1 is a flowchart of a first embodiment of a method and system for single-sample cockpit image signal target detection based on template matching;
fig. 2 is a logic block diagram of a first embodiment of a method and system for detecting a single-sample cockpit image signal target based on template matching.
Detailed Description
The following is further detailed by way of specific embodiments:
example one
The method for detecting the target of the single-sample cockpit image signal based on template matching, as shown in fig. 1, comprises the following steps:
and an edge detection step, which is to perform gray level processing on the stored template image and the cabin image acquired in real time, and then perform edge detection on the template image and the cabin image to respectively obtain a template edge image and a cabin edge image. The acquisition mode of the template image in the embodiment is as follows: 1. fixing a camera on a steering wheel of the test vehicle, and shooting an image of a cockpit instrument panel to be used as a cockpit instrument panel image; 2. since the camera is fixed, the area of the shot image is fixed, and the position of the icon in the area of the shot image is fixed, the designated area of the framing image can be set, namely, the image of the icon in the image of the instrument panel of the shot cockpit is framed according to the designated area, and the image is the template image.
The edge detection step specifically comprises an image preprocessing substep and an edge detection substep, and is used for weakening the influence of illumination on the acquired cockpit image of the cockpit instrument panel in the driving process of the automobile on the road, and specifically comprises the following steps:
and the image preprocessing substep is used for carrying out image standardization processing, gray level processing and noise reduction processing on the stored template image and the cabin image acquired in real time. In this embodiment, the preferred noise reduction processing mode is gaussian blur, specifically, a gaussian kernel with a size of 3 is adopted;
and an edge detection substep, which is used for carrying out edge detection on the template image and the cockpit image which are subjected to image standardization processing and noise reduction processing to respectively obtain a template edge image and a cockpit edge image. In the present embodiment, Canny edge detection technology is preferably used to perform edge detection on the cabin image and the template image.
Taking Canny edge detection on a cabin image as an example, the Canny edge detection technology sequentially comprises the steps of adopting a Rotot operator to calculate the gradient of the cabin image subjected to Gaussian blur processing, and in other embodiments, operators with selectable gradient calculation also comprise the Rotot operator, a Sobel operator, a Prewitt operator and the like; calculating the edge amplitude and angle of the cockpit image according to the gradient; carrying out non-maximum signal suppression on the cockpit image, thinning edge pixels of the cockpit image and realizing edge thinning; adopting double thresholds to accept or reject edge pixels to realize edge connection; and carrying out binarization processing on the cabin image and outputting a cabin edge image.
The edge amplitude and angle calculation formula is:
suppose the horizontal gradient G of pixel points (x, y) in the cockpit imagex(x, y) gradient in vertical direction Gy(x, y) and pixel value H (x, y) are:
Gx(x,y)＝H(x+1,y)-H(x-1,y)
Gy(x,y)＝H(x,y+1)-H(x,y-1)
the edge amplitude G (x, y) and the angle α (x, y) at the pixel point (x, y) are:
and a template matching step, namely detecting a matched part in the cockpit edge image as a target area, specifically detecting the cockpit edge image. Therefore, the subsequent calculation range of the color mean value in the cockpit edge image is narrowed, the calculation amount is reduced, and the timeliness is better.
And a state judgment step, namely performing color restoration on the target area, calculating the color mean value of the target area after color restoration, and judging the sign state according to the color mean value, wherein the color mean value is the color mean value of a color three channel.
As shown in fig. 2, the single-sample cockpit image signal target detection system based on template matching includes an image acquisition device and an edge calculation device, where the image acquisition device is used to capture images of a cockpit instrument panel and acquire images of a cockpit of the cockpit instrument panel, and a camera is used by the image acquisition device in this embodiment. The edge computing device comprises a template generating module, an edge detecting module, a template matching module and a state judging module, wherein:
the template generation module is used for performing frame selection on the images of the icons in the shot cockpit instrument panel images according to the designated area to obtain template images.
The edge detection module is used for carrying out gray processing on the stored template image and the cabin image acquired in real time, and then carrying out edge detection on the template image and the cabin image to respectively obtain a template edge image and a cabin edge image. Specifically, the edge detection module includes a preprocessing sub-module and an edge detection sub-module:
the preprocessing submodule is used for carrying out image standardization processing, gray level processing and noise reduction processing on the stored template image and the cabin image acquired in real time; in this embodiment, the preprocessing sub-module preferably performs noise reduction on the received template image and cockpit image in a gaussian fuzzy noise reduction mode, specifically, a gaussian kernel with a size of 3;
the edge detection submodule is used for carrying out edge detection on the template image and the cockpit image which are processed by the preprocessing submodule to respectively obtain a template edge image and a cockpit edge image. In this embodiment, the edge detection module preferably performs edge detection on the template image and the cockpit image processed by the preprocessing sub-module by using a Canny edge detection technology.
The template matching module is used for capturing a part matched with the template edge image in the cockpit edge image as a target area;
the state judgment module is used for carrying out color restoration on the target area, calculating the color mean value of the target area after color restoration, and then judging the sign state according to the color mean value, wherein the color mean value is the color mean value of the color three channels.
The specific implementation process comprises the following steps: firstly, a template image is obtained, a camera fixed on a test vehicle is used for shooting a cockpit instrument panel image, and then a template generation module selects an image of an icon in the shot cockpit instrument panel image according to a specified area frame to be used as the template image and stores the template image.
In the actual driving process, cabin images acquired by the camera on the vehicle in real time are sent to the edge computing equipment. And a preprocessing submodule in the edge computing equipment is used for carrying out image standardization processing and gray level processing on the stored template image and the received cockpit image, and carrying out noise reduction processing on the stored template image and the received cockpit image by adopting Gaussian blur of a Gaussian kernel with the size of 3, so that the influence of illumination on the image is removed, and the robustness of the image is improved. And then, the edge detection submodule carries out edge detection on the template image and the cockpit image which are processed by the preprocessing submodule by adopting a Canny edge detection technology, reduces the search range of subsequent matching and the calculation range of the color mean value, and respectively obtains the template edge image and the cockpit edge image. And then capturing a part matched with the template edge image in the cabin edge image as a target area through a template matching module. And finally, performing color restoration on the target area through a state judgment module, restoring the gray level image into a color image, calculating the color mean value of the target area with three color channels after color restoration, and judging the sign state according to whether the calculated color mean value is close to the color mean value of the edge image of the on-state template after color restoration or the color mean value of the off-state template after color restoration.
Example two
The difference from the first embodiment is that: in the template matching step, a CPU is adopted to calculate matched parts of a plurality of template edge images in the cockpit edge image in parallel as target areas.
The template matching module captures a portion matching the plurality of template edge images in the cockpit edge image as a target area by using a CPU in parallel calculation.
The cockpit instrument panel comprises a plurality of icons which are matched in sequence, the calculation speed is low, most of CPUs of modern computers have a plurality of cores, so that the template edge images of multiple targets are matched in a CPU parallel mode, the hardware advantages can be utilized, the matching calculation tasks of different template edge images are distributed to different cores to be performed synchronously, and the calculation speed is improved.
The foregoing is merely an example of the present invention and common general knowledge of known specific structures and features of the embodiments is not described herein in any greater detail. It should be noted that, for those skilled in the art, without departing from the structure of the present invention, several changes and modifications can be made, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicability of the patent. The scope of the claims of the present application shall be determined by the contents of the claims, and the description of the embodiments and the like in the specification shall be used to explain the contents of the claims.
Claims (10)
1. The method for detecting the single-sample cockpit image signal target based on template matching is characterized by comprising the following steps of:
an edge detection step, which is to perform gray level processing on the stored template image and the cabin image acquired in real time, and then perform edge detection on the template image and the cabin image to respectively obtain a template edge image and a cabin edge image;
a template matching step, namely detecting a matched part in the cockpit edge image as a target area;
and a state judgment step, namely performing color restoration on the target area, calculating the color mean value of the target area after color restoration, and judging the sign state according to the color mean value, wherein the color mean value is the color mean value of a color three channel.
2. The template matching-based single-sample cockpit image signal target detection method of claim 1, wherein: the edge detection step comprises an image pre-processing sub-step and an edge detection sub-step,
the image preprocessing substep is used for carrying out image standardization processing, gray level processing and noise reduction processing on the stored template image and the cabin image acquired in real time;
and an edge detection substep, which is used for carrying out edge detection on the template image and the cockpit image which are subjected to image standardization processing and noise reduction processing to respectively obtain a template edge image and a cockpit edge image.
3. The single-sample cockpit image signal object detection method based on template matching of claim 2, wherein: in the sub-step of edge detection, Canny edge detection technology is adopted to carry out edge detection on the cabin image and the template image.
4. The template matching-based single-sample cockpit image signal target detection method of claim 1, wherein: in the template matching step, a CPU is adopted to calculate matched parts of a plurality of template edge images in the cockpit edge image in parallel as target areas.
5. The single-sample cockpit image signal object detection method based on template matching of claim 2, wherein: and in the image preprocessing substep, Gaussian blur is adopted for noise reduction processing.
6. Single sample passenger cabin image signal target detection system based on template matching, characterized by, including edge computing device, edge computing device includes edge detection module, template matching module and state judgement module, wherein:
the edge detection module is used for carrying out gray processing on the stored template image and the cabin image acquired in real time, and then carrying out edge detection on the template image and the cabin image to respectively obtain a template edge image and a cabin edge image;
the template matching module is used for capturing a part matched with the template edge image in the cockpit edge image as a target area;
the state judgment module is used for carrying out color restoration on the target area, calculating the color mean value of the target area after color restoration, and then judging the sign state according to the color mean value, wherein the color mean value is the color mean value of the color three channels.
7. The template matching based single-sample cockpit image signal object detection system of claim 6, where the edge detection module comprises a pre-processing sub-module and an edge detection sub-module,
the preprocessing submodule is used for carrying out image standardization processing, gray level processing and noise reduction processing on the stored template image and the cabin image acquired in real time;
the edge detection submodule is used for carrying out edge detection on the template image and the cockpit image which are processed by the preprocessing submodule to respectively obtain a template edge image and a cockpit edge image.
8. The template matching-based single-sample cockpit image signal object detection system of claim 7, wherein: and the edge detection submodule carries out edge detection on the stored template image and the cabin image acquired in real time by adopting a Canny edge detection technology.
9. The template matching-based single-sample cockpit image signal object detection system of claim 6, wherein: the template matching module captures a portion matching the plurality of template edge images in the cockpit edge image as a target area by using a CPU in parallel calculation.
10. The template matching-based single-sample cockpit image signal object detection system of claim 7, wherein: and the preprocessing submodule performs noise reduction on the received template image and the cabin image by adopting Gaussian blur.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010752199.9A CN111860426A (en) | 2020-07-30 | 2020-07-30 | Single-sample cockpit image signal target detection method and system based on template matching |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010752199.9A CN111860426A (en) | 2020-07-30 | 2020-07-30 | Single-sample cockpit image signal target detection method and system based on template matching |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111860426A true CN111860426A (en) | 2020-10-30 |
Family
ID=72945721
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010752199.9A Pending CN111860426A (en) | 2020-07-30 | 2020-07-30 | Single-sample cockpit image signal target detection method and system based on template matching |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111860426A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113901884A (en) * | 2021-09-15 | 2022-01-07 | 杭州欣禾圣世科技有限公司 | Face pose estimation method, system, device and storage medium based on feature matching |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002024817A (en) * | 2000-07-10 | 2002-01-25 | Sharp Corp | Image processing method and image processing device |
| US20070132572A1 (en) * | 2004-02-20 | 2007-06-14 | Sharp Kabushiki Kaisha | Instrument panel image display device, instrument panel image changing method, vehicle, server, instrument panel image changing system, instrument panel image display program, computer-readable storage medium storing instrument panel image display program |
| US20140193099A1 (en) * | 2011-08-03 | 2014-07-10 | Hitachi Medical Corporation | Image diagnostic device and image correction method |
| CN106960207A (en) * | 2017-04-26 | 2017-07-18 | 佛山市南海区广工大数控装备协同创新研究院 | A kind of car steering position gauge field multipointer instrument automatic recognition system and method based on template matches |
| CN107767369A (en) * | 2017-09-27 | 2018-03-06 | 杭州迈锐钶科技有限公司 | A kind of the defects of buret detection method and device |
| CN109308447A (en) * | 2018-07-29 | 2019-02-05 | 国网上海市电力公司 | The method of equipment operating parameter and operating status is automatically extracted in remote monitoriong of electric power |
| CN110458093A (en) * | 2019-08-09 | 2019-11-15 | 径卫视觉科技（上海）有限公司 | A kind of Safe belt detection method and corresponding equipment based on driver's monitoring system |
-
2020
- 2020-07-30 CN CN202010752199.9A patent/CN111860426A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP2002024817A (en) * | 2000-07-10 | 2002-01-25 | Sharp Corp | Image processing method and image processing device |
| US20070132572A1 (en) * | 2004-02-20 | 2007-06-14 | Sharp Kabushiki Kaisha | Instrument panel image display device, instrument panel image changing method, vehicle, server, instrument panel image changing system, instrument panel image display program, computer-readable storage medium storing instrument panel image display program |
| US20140193099A1 (en) * | 2011-08-03 | 2014-07-10 | Hitachi Medical Corporation | Image diagnostic device and image correction method |
| CN106960207A (en) * | 2017-04-26 | 2017-07-18 | 佛山市南海区广工大数控装备协同创新研究院 | A kind of car steering position gauge field multipointer instrument automatic recognition system and method based on template matches |
| CN107767369A (en) * | 2017-09-27 | 2018-03-06 | 杭州迈锐钶科技有限公司 | A kind of the defects of buret detection method and device |
| CN109308447A (en) * | 2018-07-29 | 2019-02-05 | 国网上海市电力公司 | The method of equipment operating parameter and operating status is automatically extracted in remote monitoriong of electric power |
| CN110458093A (en) * | 2019-08-09 | 2019-11-15 | 径卫视觉科技（上海）有限公司 | A kind of Safe belt detection method and corresponding equipment based on driver's monitoring system |
Non-Patent Citations (1)
| Title |
|---|
| 金旭晖;: "基于区域颜色分割的交通标志检测和识别", 电气自动化, no. 03, 30 May 2016 (2016-05-30), pages 17 - 19 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113901884A (en) * | 2021-09-15 | 2022-01-07 | 杭州欣禾圣世科技有限公司 | Face pose estimation method, system, device and storage medium based on feature matching |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Wu et al. | Lane-mark extraction for automobiles under complex conditions | |
| CN106683119B (en) | Moving vehicle detection method based on aerial video image | |
| CN107305635A (en) | Object identifying method, object recognition equipment and classifier training method | |
| Islam et al. | Automatic vehicle number plate recognition using structured elements | |
| Farag | Real-time detection of road lane-lines for autonomous driving | |
| AG | Development of portable automatic number plate recognition (ANPR) system on Raspberry Pi | |
| CN107704833A (en) | A kind of front vehicles detection and tracking based on machine learning | |
| Su et al. | A new local-main-gradient-orientation HOG and contour differences based algorithm for object classification | |
| CN112598743A (en) | Pose estimation method of monocular visual image and related device | |
| Ashoori-Lalimi et al. | An efficient method for vehicle license plate detection in complex scenes | |
| Bulugu | Algorithm for license plate localization and recognition for tanzania car plate numbers | |
| Khinchi et al. | A review on automatic number plate recognition technology and methods | |
| Soon et al. | Malaysian car number plate detection and recognition system | |
| Khan et al. | Comparison of various edge detection filters for ANPR | |
| CN111860426A (en) | Single-sample cockpit image signal target detection method and system based on template matching | |
| Lin et al. | Two-Stage License Plate Recognition System Using Deep learning | |
| CN103093481B (en) | A kind of based on moving target detecting method under the static background of watershed segmentation | |
| Gao et al. | Vehicle detection based on color and edge information | |
| Rahaman et al. | Lane detection for autonomous vehicle management: PHT approach | |
| Khin et al. | License plate detection of Myanmar vehicle images captured from the dissimilar environmental conditions | |
| Mouzinho et al. | Hierarchical semantic segmentation based approach for road surface damages and markings detection on paved road | |
| Kodwani et al. | Automatic license plate recognition in real time videos using visual surveillance techniques | |
| Danilescu et al. | Road anomalies detection using basic morphological algorithms | |
| Kaur et al. | An Efficient Method of Number Plate Extraction from Indian Vehicles Image | |
| Satish et al. | Edge assisted fast binarization scheme for improved vehicle license plate recognition |
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 |