CN111488889B - Intelligent image processor for extracting image edges - Google Patents
Intelligent image processor for extracting image edges Download PDFInfo
- Publication number
- CN111488889B CN111488889B CN202010292008.5A CN202010292008A CN111488889B CN 111488889 B CN111488889 B CN 111488889B CN 202010292008 A CN202010292008 A CN 202010292008A CN 111488889 B CN111488889 B CN 111488889B
- Authority
- CN
- China
- Prior art keywords
- image
- module
- information
- output end
- electrically connected
- 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.)
- Active
Links
- 238000000605 extraction Methods 0.000 claims abstract description 62
- 238000004458 analytical method Methods 0.000 claims abstract description 24
- 238000012545 processing Methods 0.000 claims abstract description 23
- 239000011159 matrix material Substances 0.000 claims abstract description 17
- 238000013528 artificial neural network Methods 0.000 claims abstract description 12
- 238000006243 chemical reaction Methods 0.000 claims abstract description 12
- 230000008859 change Effects 0.000 claims description 29
- 238000010801 machine learning Methods 0.000 claims description 25
- 238000003708 edge detection Methods 0.000 claims description 19
- 230000001149 cognitive effect Effects 0.000 claims description 17
- 238000000034 method Methods 0.000 claims description 16
- 230000005540 biological transmission Effects 0.000 claims description 13
- 230000008569 process Effects 0.000 claims description 13
- 239000000284 extract Substances 0.000 claims description 6
- 230000008054 signal transmission Effects 0.000 claims description 3
- 238000005457 optimization Methods 0.000 abstract description 10
- 238000013473 artificial intelligence Methods 0.000 abstract description 5
- 238000010586 diagram Methods 0.000 description 5
- 230000019771 cognition Effects 0.000 description 3
- 230000007547 defect Effects 0.000 description 3
- 230000009471 action Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 206010063385 Intellectualisation Diseases 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 238000010191 image analysis Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 230000000007 visual effect Effects 0.000 description 1
Classifications
-
- 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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/06—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons
- G06N3/063—Physical realisation, i.e. hardware implementation of neural networks, neurons or parts of neurons using electronic means
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- General Physics & Mathematics (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Biophysics (AREA)
- Biomedical Technology (AREA)
- Mathematical Physics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Molecular Biology (AREA)
- General Health & Medical Sciences (AREA)
- Computational Linguistics (AREA)
- Multimedia (AREA)
- Neurology (AREA)
- Medical Informatics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The application discloses an intelligent image processor for extracting image edges, which particularly relates to the field of image processing, and comprises an image set acquisition module and an FPGA chip, wherein the output end of the image set acquisition module is electrically connected with an AD conversion circuit, the output end of the AD conversion circuit is electrically connected with a matrix information extractor, the output end of the matrix information extractor is electrically connected with a feature extraction module and original image data output, and an image information analysis module, a deep neural network, a comparison module and image classification output are arranged in the feature extraction module. According to the application, the artificial intelligence is set to perform early image classification, the images to be extracted are classified in a large quantity, the image information is dataized, the image recognition and the image optimization are not needed by an image edge extraction algorithm, and the FPGA chip is used as a processing circuit, so that the positioning precision of the images can be improved, the data volume in the subsequent processing of the images is reduced, and the effective and rapid operation of the image edge extraction is facilitated.
Description
Technical Field
The application relates to the technical field of image processing, in particular to an intelligent image processor for image edge extraction.
Background
The human cognition objective world and the mutual communication of people and people, the main media through the cognition objective world are images, the images are the main sources of information acquired by people, scientific data show that the information quantity acquired by people through visual images is about 4 times of the information quantity acquired by other paths, the edges of the images are the most basic characteristics of the whole images, a large amount of important information of the images is contained, the characteristics of the images at low level are defined by the edges: the edge extraction is considered to be to preserve the region of the image where the gray level change is intense, where the gray level change is greatest (where the gray level value of the image changes most severely), because edge extraction is considered to preserve the high frequency signal, which is the most intuitive method mathematically, namely differentiation (differential for digital images), and also can be said to use a high pass filter from the point of view of signal processing.
Edge information is extremely important information in an image, and theoretically, all information in the image can be recovered through the edge information, so edge detection is important content of image analysis; is critical to deal with many problems; the traditional edge detection mainly uses a difference operator in the horizontal direction and the vertical direction to detect the edges in the horizontal direction and the vertical direction respectively, then synthesizes a certain gradient to detect the edges, only needs to generate the difference in the two directions when the computer is realized, and then synthesizes the edges, but the method has obvious defects that only the edge information in the horizontal direction and the vertical direction is emphasized, but the general actual image contains edge information in a plurality of directions even in any direction, different image edge detection algorithms are applicable to different images, and the common image edge detection algorithms comprise an edge detection algorithm based on mathematical morphology, an edge detection algorithm based on graph theory and an edge detection algorithm based on multi-color characteristics, and the three detection algorithms are the most widely applied edge detection algorithms at present.
However, the edge detection algorithm based on mathematical morphology, the edge detection algorithm based on graph theory and the edge detection algorithm based on multi-color features which are used at present have certain limitations, the degree of intellectualization is low, when a large amount of image edge information extraction is processed, the image edge extraction algorithm must optimize images when the algorithm is executed, the incompatibility problem is more specific and easy to cause for the required specification of the images, the architecture is simple, the pre-classification processing of the images cannot be carried out, therefore, when different types of images appear, the algorithm identification needs to be carried out in advance, the edge extraction is carried out, the defects of increased operand, slow processing effect and the like are caused, and the efficient large-batch image processing cannot be realized, so that the further optimization is needed.
It is therefore desirable to provide an intelligent image processor for image edge extraction.
Disclosure of Invention
In order to overcome the defects in the prior art, the embodiment of the application provides an intelligent image processor for image edge extraction, which is provided with an artificial intelligence and early image classification, performs mass classification on images to be extracted, and dataizes image information, does not need an image edge extraction algorithm to perform image recognition and image optimization, and utilizes an FPGA chip as a processing circuit, so that the positioning accuracy of the images can be improved, the data volume in the subsequent processing of the images can be reduced, the effective and rapid operation of the image edge extraction is facilitated, the novel information of the images is dataized and converted through a preferential image classification system, the characteristic information content of a large number of images is extracted, the processed image-optimized data information can be directly used for image edge extraction when the image edge extraction is performed, the algorithm step of the image edge extraction is simplified, and the incompatibility problem of the image edge extraction algorithm and the image optimization is avoided, so as to solve the problem in the background technology.
In order to achieve the above purpose, the present application provides the following technical solutions: the intelligent image processor comprises an image set acquisition module and an FPGA chip, wherein an AD conversion circuit is electrically connected to the output end of the image set acquisition module, a matrix information extractor is electrically connected to the output end of the AD conversion circuit, a feature extraction module and an original image data output are electrically connected to the output end of the matrix information extractor, an image information analysis module and a deep neural network are arranged in the feature extraction module, a comparison module is electrically connected to the output end of the feature extraction module, an image classification output is electrically connected to the output end of the comparison module, an original image data output and an image classification output are electrically connected to the input end of the FPGA chip, an image edge extraction module is arranged in the FPGA chip, and the image set acquisition module comprises a camera assembly, a storage and an image set transmission module.
In a preferred embodiment, the image set acquisition module includes: after the scanning strategy acquires the first image information in real time, the resolution strategy extracts the definition value of the first image and compares the definition value with the image definition threshold, and when the definition value of the first image is lower than the image definition threshold, the image processing unit outputs the re-shooting information; otherwise, numbering the first image, and sending second image extraction information to the camera component; after the scanning strategy acquires the second image information in real time, the resolution strategy extracts the definition value of the second image and compares the definition value with the image definition threshold, and when the definition value of the second image is lower than the image definition threshold, the re-shooting information is output; and otherwise, numbering the second image, and comparing the sharpness values of the numbered first image and the second image to output the current image information with the highest sharpness value.
In a preferred embodiment, the storage includes a random storage and a Flash storage, the image set transmission module is a wired or wireless signal transmission device, and an input end of the image set transmission module is electrically connected with an output end of the random storage.
In a preferred embodiment, the image information analysis module includes a viewpoint change process for identifying whether an object is photographed or acquired resulting in an image that is original or multi-dimensionally rotated, a zoom change for identifying whether it is an image of the same object that is different in size from the object, an intra change for identifying all correct categories of all the same photographed object for category discrimination, and a joint process for applying one or more of the viewpoint change process, the zoom change, and the intra change to the collaborative identification analysis.
In a preferred embodiment, the deep neural network is internally provided with machine learning comprising a traditional logic study consisting of an auxiliary machine learning model, a cognitive model consisting of a target machine learning model, and a theoretical analysis consisting of a comparison component, the traditional logic study and the cognitive model being from different machine learning model classes, the cognitive model being a limited capacity machine learning model.
In a preferred embodiment, the conventional logic study is configured to assign a first score to an unlabeled observation, the cognitive model is configured to assign a second score to the unlabeled observation, the theoretical analysis is configured to compare the first score and the second score to determine a probability that the cognitive model has returned a false positive or false negative result, the comparing component of the first score and the second score is further configured to perform a comparison comprising: determining a magnitude of a difference between the first fraction and the second fraction; determining that the target machine learning model has returned a false positive when the magnitude is negative; and determining that the target machine learning model has returned a false negative when the magnitude is positive.
In a preferred embodiment, the image edge extraction module includes a restoration image circuit, an image filter circuit, an edge enhancement circuit, an edge detection circuit, and an edge positioning circuit, where an output end of the restoration image circuit is electrically connected to an output end of the image filter circuit, an output end of the image filter circuit is electrically connected to an input end of the edge enhancement circuit, and an input end and an output end of the edge detection circuit are respectively electrically connected to an output end of the edge enhancement circuit and an input end of the edge positioning circuit.
In a preferred embodiment, the input end of the FPGA chip is electrically connected with an image information sorting module, the image information sorting module is used for corresponding the image classification information output by the original image data and the original image data information output by the image classification, and the input end of the image restoration circuit is electrically connected with the output end of the matrix information extractor.
The application has the technical effects and advantages that:
1. according to the application, the image to be extracted is classified in a large quantity by setting the artificial intelligence to be careful and the image information is dataized, the image recognition and the image optimization are not needed by an image edge extraction algorithm, and the FPGA chip is used as a processing circuit, so that the positioning precision of the image can be improved, the data volume in the subsequent processing of the image is reduced, and the effective and rapid operation of the image edge extraction is facilitated;
2. according to the application, through an artificial intelligence mode, the algorithm initially builds a traditional logic research and cognition model through the existing artificial analysis and establishment logic, records the difference between prediction and correct output through a machine learning mode, tunes the input weight to improve the prediction accuracy, and when the image classification system is used for a long time, the image classification system is continuously optimized, so that the image optimization and classification intelligent accuracy is gradually improved.
3. According to the application, the novel image information is subjected to data conversion through the preferential image classification system, and the characteristic information content of a large number of images is extracted, so that the processed image-optimized data information can be directly utilized for image edge extraction during image edge extraction, the algorithm steps of image edge extraction are simplified, and the problem of incompatibility between an image edge extraction algorithm and image optimization is avoided.
Drawings
Fig. 1 is a schematic diagram of the overall structure of the present application.
Fig. 2 is a schematic diagram of an image set acquisition module according to the present application.
Fig. 3 is a schematic diagram of an image edge extraction module according to the present application.
Fig. 4 is a schematic diagram of an image information analysis module according to the present application.
Fig. 5 is a schematic diagram of a deep neural network according to the present application.
The reference numerals are: 1. an image set acquisition module; 11. a camera assembly; 12. a reservoir; 13. an image set transmission module; 2. an AD conversion circuit; 3. a matrix information extractor; 4. a feature extraction module; 41. an image information analysis module; 411. performing viewpoint change processing; 412. scaling the change; 413. an internal variation; 414. joint treatment; 42. a deep neural network; 421. machine learning; 5. outputting original image data; 6. comparison module; 7. classifying and outputting images; 8. an FPGA chip; 9. an image edge extraction module; 91. a reduction image circuit; 92. an image filter circuit; 93. an edge enhancement circuit; 94. an edge detection circuit; 95. edge locating circuitry.
Detailed Description
The following description of the embodiments of the present application will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present application, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the application without making any inventive effort, are intended to be within the scope of the application.
An intelligent image processor for image edge extraction as shown in fig. 1-5 comprises an image set acquisition module 1 and an FPGA chip 8, wherein the output end of the image set acquisition module 1 is electrically connected with an AD conversion circuit 2, the output end of the AD conversion circuit 2 is electrically connected with a matrix information extractor 3, the output end of the matrix information extractor 3 is electrically connected with a feature extraction module 4 and an original image data output 5, an image information analysis module 41 and a deep neural network 42 are arranged in the feature extraction module 4, the output end of the feature extraction module 4 is electrically connected with a comparison module 6, the output end of the comparison module 6 is electrically connected with an image classification output 7, the output end of the original image data output 5 and the image classification output 7 are electrically connected with the input end of the FPGA chip 8, the interior of the FPGA chip 8 is provided with an image edge extraction module 9, the image set acquisition module 1 comprises a camera assembly 11, a storage 12 and an image set transmission module 13, and the output end of the camera assembly 11 is electrically connected with the input ends of the storage 12 and the image set transmission module 13.
The implementation mode specifically comprises the following steps: the method has the advantages that the artificial intelligence is set to be careful in pre-image classification, a large amount of classification is carried out on images to be extracted, image information is dataized, image recognition and image optimization are carried out without an image edge extraction algorithm, an FPGA chip is used as a processing circuit, the positioning precision of the images can be improved, meanwhile, the data volume in the subsequent processing of the images is reduced, the effective and rapid operation of image edge extraction is facilitated, the data conversion is carried out on novel information of the images through a preferential image classification system, the characteristic information content of a large number of images is extracted, the image edge extraction can be carried out by directly utilizing the processed image optimized data information when the image edge extraction is carried out, the algorithm step of the image edge extraction is simplified, and the incompatibility problem of the image edge extraction algorithm and the image optimization is avoided.
The image set acquisition module 1 includes: after the scanning strategy acquires the first image information in real time, the resolution strategy extracts the definition value of the first image and compares the definition value with the image definition threshold, and when the definition value of the first image is lower than the image definition threshold, the image processing unit outputs the re-shooting information; otherwise, numbering the first image, and sending second image extraction information to the camera component 11; after the scanning strategy acquires the second image information in real time, the resolution strategy extracts the definition value of the second image and compares the definition value with the image definition threshold, and when the definition value of the second image is lower than the image definition threshold, the re-shooting information is output; otherwise, numbering the second image, comparing the sharpness values of the numbered first image and the second image, and outputting the current image information with the highest sharpness value, wherein the high-sharpness image is convenient for subsequent image processing.
The storage 12 includes a random storage and a Flash storage, the image set transmission module 13 is a wired or wireless signal transmission device, and an input end of the image set transmission module 13 is electrically connected with an output end of the random storage, and is used for sorting and transmitting image information of the image set acquisition module 1.
The image information analysis module 41 includes a viewpoint change process 411, a scaling change 412, an intra-change 413, and a joint process 414, where the viewpoint change process 411 is used to identify whether an object is photographed or acquired and what is an image of the original or multi-dimensional rotation, the scaling change 412 is used to identify whether the object is an image of the same object and different in size, the intra-change 413 is used to identify all correct types of all the same photographed object for type discrimination, and the joint process 414 is used to apply one or more of the viewpoint change process 411, the scaling change 412, and the intra-change 413 to a collaborative identification analysis, and the images are classified and identified by using multiple kinds of change analysis.
The deep neural network 42 is internally provided with a machine learning 421, the machine learning 421 comprises a traditional logic study, a cognitive model and a theoretical analysis, the traditional logic study is composed of an auxiliary machine learning model, the cognitive model is composed of a target machine learning model, the traditional logic study and the cognitive model come from different machine learning model types, the cognitive model is a machine learning model with limited capacity, the theoretical analysis is composed of a comparison component, the deep neural network 42 is built, the intelligent image classification is realized by using the deep neural network 42, and the accuracy is improved.
Wherein the conventional logic study is configured to assign a first score to the unlabeled observations, the cognitive model is configured to assign a second score to the unlabeled observations, the theoretical analysis is configured to compare the first score to the second score to determine a probability that the cognitive model has returned a false positive or false negative result, the comparing component of the first score and the second score is further configured to perform a comparison comprising: determining a magnitude of a difference between the first fraction and the second fraction; determining that the target machine learning model has returned a false positive when the magnitude is negative; and when the amplitude is positive, determining that the target machine learning model returns false negatives, and realizing automatic optimization and upgrading of the intelligent image recognition and classification system to be intelligent.
The image edge extraction module 9 includes a restoring image circuit 91, an image filter circuit 92, an edge enhancement circuit 93, an edge detection circuit 94 and an edge positioning circuit 95, wherein an output end of the restoring image circuit 91 is electrically connected with an output end of the image filter circuit 92, an output end of the image filter circuit 92 is electrically connected with an input end of the edge enhancement circuit 93, and an input end and an output end of the edge detection circuit 94 are respectively electrically connected with an output end of the edge enhancement circuit 93 and an input end of the edge positioning circuit 95, and the whole image edge of the component is detected with an extraction program.
The input end of the FPGA chip 8 is electrically connected with an image information sorting module, which is used for corresponding the image classification information of the original image data output 5 and the original image data information of the image classification output 7, and the input end of the image restoration circuit 91 is electrically connected with the output end of the matrix information extractor 3, so that the processed image-optimized data information is directly used for image edge extraction, and the algorithm steps of image edge extraction are simplified.
The model of the camera assembly 11 is of the SONY IMX362 type.
The working principle of the application is as follows:
the first step: after the camera component 11 is used for acquiring images, the image set transmission module 13 is used for transmitting a large number of images stored in the storage 12 to the AD conversion circuit 2 in a data transmission mode, the AD conversion circuit 2 is used for converting the images into gray images of 28 x 28 so as to form a pixel matrix, the image information is converted into a data signal form so as to facilitate the algorithm to recognize and analyze the image information, and then the matrix information extractor 3 is used for extracting the information of the pixel matrix;
and a second step of: the matrix information extractor 3 respectively transmits the extracted image matrix information to the feature extraction module 4 and the FPGA chip 8, the image matrix information is processed by utilizing the image information analysis module 41 to perform viewpoint change processing 411, scaling change 412, internal change 413 and joint processing 414 under the variable intervention of the deep neural network 42 at the feature extraction module 4, the content of the image is quantized, and then the quantized data of each image is compared under the action of the comparison module 6, similarity analysis is performed, and a classification output value FPGA chip 8 is obtained;
and a third step of: then, under the action of the FPGA chip 8, image edge information is extracted through a series of circuits including a restored image circuit 91, an image filtering circuit 92, an edge enhancement circuit 93, an edge detection circuit 94 and an edge positioning circuit 95, and the FPGA chip is used as a processing circuit, so that the positioning precision of an image can be improved, the data volume in the subsequent processing of the image is reduced, and the effective and rapid operation of image edge extraction is facilitated.
The last points to be described are: first, in the description of the present application, it should be noted that, unless otherwise specified and defined, the terms "mounted," "connected," and "connected" are to be construed broadly, and may be mechanical or electrical, or may be a direct connection between two elements, and "upper," "lower," "left," "right," etc. are merely used to indicate relative positional relationships, which may be changed when the absolute position of the object being described is changed;
secondly: in the drawings of the disclosed embodiments, only the structures related to the embodiments of the present disclosure are referred to, and other structures can refer to the common design, so that the same embodiment and different embodiments of the present disclosure can be combined with each other under the condition of no conflict;
finally: the foregoing description of the preferred embodiments of the application is not intended to limit the application to the precise form disclosed, and any such modifications, equivalents, and alternatives falling within the spirit and principles of the application are intended to be included within the scope of the application.
Claims (5)
1. An intelligent image processor for image edge extraction, which comprises an image set acquisition module (1) and an FPGA chip (8), and is characterized in that: the device comprises an image collection acquisition module (1), an AD conversion circuit (2), a matrix information extractor (3), a feature extraction module (4) and an original image data output (5), wherein the image information analysis module (41) and a deep neural network (42) are arranged in the feature extraction module (4), the output end of the feature extraction module (4) is electrically connected with a comparison module (6), the output end of the comparison module (6) is electrically connected with an image classification output (7), the output end of the original image data output (5) and the output end of the image classification output (7) are electrically connected with the input end of an FPGA chip (8), an image edge extraction module (9) is arranged in the FPGA chip (8), and the image collection acquisition module (1) comprises a camera assembly (11), a storage (12) and an image collection transmission module (13), and the output end of the camera assembly (11) is electrically connected with the input end of the storage (12) and the image collection module (13);
the image set acquisition module (1) comprises: after the scanning strategy acquires the first image information in real time, the resolution strategy extracts the definition value of the first image and compares the definition value with the image definition threshold, and when the definition value of the first image is lower than the image definition threshold, the image processing unit outputs the re-shooting information; otherwise, numbering the first image, and sending second image extraction information to the camera component (11); after the scanning strategy acquires the second image information in real time, the resolution strategy extracts the definition value of the second image and compares the definition value with the image definition threshold, and when the definition value of the second image is lower than the image definition threshold, the re-shooting information is output; otherwise, numbering the second image, and comparing the sharpness values of the numbered first image and the second image to output the current image information with the highest sharpness value;
the image information analysis module (41) comprises a viewpoint change process (411), a scaling change (412), an internal change (413) and a joint process (414), wherein the viewpoint change process (411) is used for identifying whether an object is photographed or acquired and an original or multi-dimensional rotated image is caused, the scaling change (412) is used for identifying whether the object is an image with the same object and different size, the internal change (413) is used for identifying all correct types of all the same photographed objects to conduct category distinction, and the joint process (414) is used for applying one or more of the viewpoint change process (411), the scaling change (412) and the internal change (413) to collaborative identification analysis;
the deep neural network (42) is internally provided with machine learning (421), the machine learning (421) comprises a traditional logic study, a cognitive model and a theoretical analysis, the traditional logic study is composed of an auxiliary machine learning model, the cognitive model is composed of a target machine learning model, the traditional logic study and the cognitive model come from different machine learning model categories, the cognitive model is a limited-capacity machine learning model, and the theoretical analysis is composed of a comparison component.
2. An intelligent image processor for image edge extraction as defined in claim 1, wherein: the storage (12) comprises a random storage and a Flash storage, the image set transmission module (13) is wired or wireless signal transmission equipment, and the input end of the image set transmission module (13) is electrically connected with the output end of the random storage.
3. An intelligent image processor for image edge extraction according to claim 2, wherein: the conventional logic study is configured to assign a first score to an unlabeled observation, the cognitive model is configured to assign a second score to the unlabeled observation, the theoretical analysis is configured to compare the first score and the second score to determine a probability that the cognitive model has returned a false positive or false negative result, the comparing component of the first score and the second score is further configured to perform a comparison comprising: determining a magnitude of a difference between the first fraction and the second fraction; determining that the target machine learning model has returned a false positive when the magnitude is negative; and determining that the target machine learning model has returned a false negative when the magnitude is positive.
4. An intelligent image processor for image edge extraction according to claim 3, wherein: the image edge extraction module (9) comprises a restored image circuit (91), an image filter circuit (92), an edge enhancement circuit (93), an edge detection circuit (94) and an edge positioning circuit (95), wherein the output end of the restored image circuit (91) is electrically connected with the output end of the image filter circuit (92), the output end of the image filter circuit (92) is electrically connected with the input end of the edge enhancement circuit (93), and the input end and the output end of the edge detection circuit (94) are electrically connected with the output end of the edge enhancement circuit (93) and the input end of the edge positioning circuit (95) respectively.
5. An intelligent image processor for image edge extraction as defined in claim 4, wherein: the input end of the FPGA chip (8) is electrically connected with an image information arrangement module, the image information arrangement module is used for corresponding the image classification information of the original image data output (5) and the original image data information of the image classification output (7), and the input end of the restored image circuit (91) is electrically connected with the output end of the matrix information extractor (3).
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010292008.5A CN111488889B (en) | 2020-04-14 | 2020-04-14 | Intelligent image processor for extracting image edges |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010292008.5A CN111488889B (en) | 2020-04-14 | 2020-04-14 | Intelligent image processor for extracting image edges |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111488889A CN111488889A (en) | 2020-08-04 |
| CN111488889B true CN111488889B (en) | 2023-11-07 |
Family
ID=71795027
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010292008.5A Active CN111488889B (en) | 2020-04-14 | 2020-04-14 | Intelligent image processor for extracting image edges |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111488889B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113178164B (en) * | 2020-10-12 | 2022-06-03 | 浙江山泓科技有限公司 | Intelligent image processing device for LED display screen |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106067026A (en) * | 2016-05-30 | 2016-11-02 | 天水师范学院 | A kind of Feature extraction and recognition search method of microimage of Chinese medical herb |
| CN106097335A (en) * | 2016-06-08 | 2016-11-09 | 安翰光电技术（武汉）有限公司 | Digestive tract focus image identification system and recognition methods |
| CN107945197A (en) * | 2017-12-20 | 2018-04-20 | 南通使爱智能科技有限公司 | A kind of intelligent image processing instrument for Edge extraction |
| CN108986125A (en) * | 2017-11-30 | 2018-12-11 | 成都通甲优博科技有限责任公司 | Object edge extracting method, device and electronic equipment |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8023699B2 (en) * | 2007-03-09 | 2011-09-20 | Jiris Co., Ltd. | Iris recognition system, a method thereof, and an encryption system using the same |
| US10740655B2 (en) * | 2018-07-02 | 2020-08-11 | Centre Hospitalier Universitaire Vaudois | Integrative prediction of a cognitive evolution of a subject |
-
2020
- 2020-04-14 CN CN202010292008.5A patent/CN111488889B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106067026A (en) * | 2016-05-30 | 2016-11-02 | 天水师范学院 | A kind of Feature extraction and recognition search method of microimage of Chinese medical herb |
| CN106097335A (en) * | 2016-06-08 | 2016-11-09 | 安翰光电技术（武汉）有限公司 | Digestive tract focus image identification system and recognition methods |
| CN108986125A (en) * | 2017-11-30 | 2018-12-11 | 成都通甲优博科技有限责任公司 | Object edge extracting method, device and electronic equipment |
| CN107945197A (en) * | 2017-12-20 | 2018-04-20 | 南通使爱智能科技有限公司 | A kind of intelligent image processing instrument for Edge extraction |
Non-Patent Citations (2)
| Title |
|---|
| 刘千 ; 葛阿雷 ; 史伟 ; .形态学与RCF相结合的唐卡图像边缘检测算法.计算机应用与软件.2019,(06),全文. * |
| 张春雷 ; .基于并行卷积神经网络的军事目标图像分类技术.电子设计工程.2019,(08),全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111488889A (en) | 2020-08-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Türkyılmaz et al. | License plate recognition system using artificial neural networks | |
| CN110765855A (en) | Pathological image processing method and system | |
| CN110335233B (en) | Highway guardrail plate defect detection system and method based on image processing technology | |
| CN109190456B (en) | Multi-feature fusion overlook pedestrian detection method based on aggregated channel features and gray level co-occurrence matrix | |
| Cun et al. | Image splicing localization via semi-global network and fully connected conditional random fields | |
| CN112257711B (en) | Method for detecting damage fault of railway wagon floor | |
| CN113160200A (en) | Industrial image defect detection method and system based on multitask twin network | |
| Rasooli et al. | Farsi license plate detection based on element analysis and characters recognition | |
| CN110008949B (en) | Image target detection method, system, device and storage medium | |
| CN116681979A (en) | Power equipment target detection method under complex environment | |
| CN111488889B (en) | Intelligent image processor for extracting image edges | |
| CN112883827B (en) | Method and device for identifying specified target in image, electronic equipment and storage medium | |
| CN111597875A (en) | Traffic sign identification method, device, equipment and storage medium | |
| CN112949451B (en) | Cross-modal target tracking method and system through modal perception feature learning | |
| CN113705361A (en) | Method and device for detecting model in living body and electronic equipment | |
| Patki et al. | A review: Cotton leaf disease detection | |
| CN110610177A (en) | Training method of character recognition model, character recognition method and device | |
| CN116912872A (en) | Drawing identification method, device, equipment and readable storage medium | |
| Saranya et al. | An approach towards ear feature extraction for human identification | |
| Truong et al. | A study on visual saliency detection in infrared images using Boolean map approach | |
| CN114582012A (en) | Skeleton human behavior recognition method, device and equipment | |
| CN114387670A (en) | Gait recognition method and device based on space-time feature fusion and storage medium | |
| CN111259981A (en) | Automatic classification system after remote sensing image processing | |
| Wang et al. | Fast loop closure detection via binary content | |
| Nguyen et al. | Fast scene text detection with RT-LoG operator and CNN |
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 |