CN117496532A - Intelligent recognition tool based on 0CR - Google Patents
Intelligent recognition tool based on 0CR Download PDFInfo
- Publication number
- CN117496532A CN117496532A CN202311487549.3A CN202311487549A CN117496532A CN 117496532 A CN117496532 A CN 117496532A CN 202311487549 A CN202311487549 A CN 202311487549A CN 117496532 A CN117496532 A CN 117496532A
- Authority
- CN
- China
- Prior art keywords
- image
- module
- features
- feature
- pixel value
- 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
- 238000007781 pre-processing Methods 0.000 claims abstract description 27
- 238000000605 extraction Methods 0.000 claims abstract description 23
- 238000000034 method Methods 0.000 claims description 29
- 238000004422 calculation algorithm Methods 0.000 claims description 27
- 238000005286 illumination Methods 0.000 claims description 26
- 230000008569 process Effects 0.000 claims description 19
- 239000011159 matrix material Substances 0.000 claims description 17
- 230000002159 abnormal effect Effects 0.000 claims description 12
- 238000012937 correction Methods 0.000 claims description 11
- 239000003086 colorant Substances 0.000 claims description 9
- 238000003709 image segmentation Methods 0.000 claims description 7
- 230000008030 elimination Effects 0.000 claims description 4
- 238000003379 elimination reaction Methods 0.000 claims description 4
- 230000015572 biosynthetic process Effects 0.000 claims description 3
- 238000013507 mapping Methods 0.000 claims description 3
- 230000009467 reduction Effects 0.000 claims description 3
- 239000013598 vector Substances 0.000 claims description 3
- 239000000284 extract Substances 0.000 claims 1
- 230000000694 effects Effects 0.000 abstract description 10
- 238000012545 processing Methods 0.000 description 12
- 238000005516 engineering process Methods 0.000 description 8
- 238000009877 rendering Methods 0.000 description 5
- 238000004364 calculation method Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 238000012986 modification Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000010606 normalization Methods 0.000 description 3
- 230000009466 transformation Effects 0.000 description 3
- 206010063385 Intellectualisation Diseases 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 2
- 238000012958 reprocessing Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 230000004913 activation Effects 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000013135 deep learning Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000003708 edge detection Methods 0.000 description 1
- 238000001914 filtration Methods 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000005055 memory storage Effects 0.000 description 1
- 238000005457 optimization Methods 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
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/19—Recognition using electronic means
- G06V30/191—Design or setup of recognition systems or techniques; Extraction of features in feature space; Clustering techniques; Blind source separation
- G06V30/19173—Classification techniques
-
- 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
- G06N3/0464—Convolutional networks [CNN, ConvNet]
-
- 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
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/14—Image acquisition
- G06V30/1431—Illumination control
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/16—Image preprocessing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/18—Extraction of features or characteristics of the image
- G06V30/1801—Detecting partial patterns, e.g. edges or contours, or configurations, e.g. loops, corners, strokes or intersections
- G06V30/18019—Detecting partial patterns, e.g. edges or contours, or configurations, e.g. loops, corners, strokes or intersections by matching or filtering
- G06V30/18038—Biologically-inspired filters, e.g. difference of Gaussians [DoG], Gabor filters
- G06V30/18048—Biologically-inspired filters, e.g. difference of Gaussians [DoG], Gabor filters with interaction between the responses of different filters, e.g. cortical complex cells
- G06V30/18057—Integrating the filters into a hierarchical structure, e.g. convolutional neural networks [CNN]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/19—Recognition using electronic means
- G06V30/19007—Matching; Proximity measures
- G06V30/19013—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
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/19—Recognition using electronic means
- G06V30/191—Design or setup of recognition systems or techniques; Extraction of features in feature space; Clustering techniques; Blind source separation
- G06V30/1912—Selecting the most significant subset of features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/24—Character recognition characterised by the processing or recognition method
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V30/00—Character recognition; Recognising digital ink; Document-oriented image-based pattern recognition
- G06V30/10—Character recognition
- G06V30/26—Techniques for post-processing, e.g. correcting the recognition result
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Multimedia (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Software Systems (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Medical Informatics (AREA)
- Databases & Information Systems (AREA)
- Biodiversity & Conservation Biology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Image Analysis (AREA)
Abstract
The invention discloses a 0 CR-based intelligent recognition tool, which belongs to the field of image recognition tools and comprises a data acquisition module, a preprocessing module, a feature extraction and selection module, a classification decision module and an information recognition module; the data acquisition module is used for acquiring the image file to be processed transmitted by the terminal equipment; the preprocessing module is used for preprocessing the image file to be processed, wherein the preprocessing comprises the steps of increasing the definition of the image to be processed, adjusting the angle of the image and normalizing; the feature extraction and selection module is used for extracting and selecting the same feature information of the preprocessed image file; the classification decision module is used for acquiring the information of the image to be processed and classifying the image according to the scene type; the information identification module is used for matching the classification and result of the image acquisition with the original template in the image identification information base and outputting a matching identification result in the image to be processed. The intelligent recognition tool can effectively improve recognition effect and recognition speed.
Description
Technical Field
The invention relates to an image recognition tool, in particular to a smart recognition tool based on 0 CR.
Background
0CR (image recognition technology) refers to a process in which an electronic device (e.g., a scanner or a digital camera) checks characters printed on paper, determines the shape thereof by detecting dark and light patterns, and translates the shape into computer text by a character recognition method. OCR image recognition technology belongs to artificial intelligence field, and the main index of measuring the performance of an OCR system is: rejection rate, false recognition rate, recognition speed, user interface friendliness, product stability, usability, feasibility and the like.
At present, the image recognition technology on the market basically performs training and learning based on a large amount of data with high quality of original images, and cannot meet the problems of memory storage, recognition failure and the like. The following problems are mainly categorized: affected by light: under the influence of light, the light sensing capability of the camera is disturbed, and particularly, the model required at different stages in the daytime is different from the model required at night, so that the influence of random events cannot be completely solved. Limited to image effects: if the interference elements increase during image recognition due to the problems of external interference, poor image quality (such as image blurring) and the like, the accuracy of image recognition may be affected. A large amount of annotation data is needed for training: many deep learning algorithms require a large amount of annotation data to train to achieve higher recognition accuracy. However, acquiring these annotation data requires a significant amount of manpower, material resources, and time. The calculated amount is large: image recognition algorithms typically require extensive computation and high performance computer hardware support, which limits their application to large-scale data sets. Therefore, how to improve the recognition effect and recognition speed of the image recognition tool is a current urgent problem to be solved.
Accordingly, one skilled in the art would provide a 0 CR-based intelligent recognition tool to solve the problems set forth in the background.
Disclosure of Invention
The invention aims to provide a 0 CR-based intelligent recognition tool which can effectively improve recognition effect and recognition speed so as to solve the problems in the prior art.
In order to achieve the above purpose, the present invention provides the following technical solutions:
the intelligent recognition tool based on 0CR comprises a data acquisition module, a preprocessing module, a feature extraction and selection module, a classification decision module and an information recognition module;
the data acquisition module is used for acquiring an image file to be processed transmitted by the terminal equipment; the preprocessing module is used for preprocessing the image file to be processed, wherein the preprocessing comprises the steps of increasing the definition of the image to be processed, adjusting the angle of the image and normalizing; the feature extraction and selection module is used for extracting and selecting the same feature information of the preprocessed image file; the classification decision module is used for acquiring the information of the image to be processed and classifying the image according to the scene type; the information identification module is used for matching the classification and result of the image acquisition with the original template in the image identification information base and outputting a matching identification result in the image to be processed.
As a further scheme of the invention: the specific process of preprocessing the image file to be processed by the preprocessing module is as follows:
normalizing the size of the acquired image, and acquiring illumination intensity parameters related to the acquired image;
judging whether the collected image has abnormal illumination or not based on the illumination intensity parameter, and simultaneously analyzing the abnormal type of the abnormal illumination image;
carrying out illumination correction on the illumination abnormal image;
performing dotting and positioning on the image file with the increased definition of the image to be processed to obtain image features with background colors;
image segmentation is performed on the image based on the set of image background colors.
As still further aspects of the invention: the feature extraction and selection module is used for carrying out reprocessing on the image file before extracting and selecting the same feature information of the preprocessed image file, and specifically comprises the following steps:
based on an image segmentation recognition result of the illumination correction image, invoking a corresponding shielding elimination algorithm to clear shielding information of image features of the object to be recognized;
and calling a corresponding background splashing algorithm to perform background removing and subtracting operation on the image features of the objects to be identified, and forming feature vectors by utilizing the image features of the objects to be identified.
As still further aspects of the invention: the specific process of extracting and selecting the same characteristic information of the preprocessed image file by the characteristic extracting and selecting module is as follows:
(1) Feature classification: physical features, structural features, mathematical features;
(2) Feature formation generating a set of basic features from the identified image, such generated features being called raw features;
(3) Extracting features, namely representing samples by a low-dimensional space through a mapping method, wherein the mapped secondary features are linear combinations of original features;
(4) Feature selection a number of most efficient features are selected from a set of features to achieve a reduction in feature space dimension.
As still further aspects of the invention: the specific process of classifying the images according to the scene types by the classification decision module is as follows:
preprocessing an illumination correction image, specifically: determining a difference operator, and carrying out gray level difference in the neighborhood to obtain a difference gray level map; determining a color difference valve value, and separating an image area to be identified in the differential gray scale image from background colors by using the color difference valve value;
and determining a pixel value interval to which a pixel value of each background color area pixel point in the image belongs, and obtaining a first pixel identification information sequence based on a corresponding relation between the pixel value interval to which the pixel value of each pixel point belongs and preset pixel value interval identification information.
As still further aspects of the invention: the specific obtaining process of the first pixel identification information sequence comprises the following steps:
determining a plurality of continuous pixel value intervals divided in advance, and determining a pixel value interval to which each pixel value of each pixel point in the first area belongs;
dividing the first region pixel points based on the pixel value intervals to obtain at least one first pixel point pixel value string, wherein the pixel value intervals of the pixel points in any two adjacent first pixel point pixel value strings are different;
and ordering the pixel identification information of each first pixel point pixel value string to obtain a corresponding first pixel identification information sequence.
As still further aspects of the invention: the specific process of the information identification module for matching the classification and result of the image acquisition with the original template in the image identification information base is as follows:
acquiring image features of an object to be identified, and carrying out feature extraction on the image features based on a convolution algorithm to obtain an intermediate feature matrix;
sorting and reorganizing the intermediate feature matrix to obtain a post feature matrix;
performing feature extraction on the post feature matrix based on a convolution algorithm to obtain an image feature set;
and matching the image feature set with the original template in the image identification information base.
As still further aspects of the invention: and the output matching recognition result in the image to be processed is displayed through the mobile terminal and the computer terminal, and meanwhile, the manual adjustment of the recognition result is supported, so that the accuracy of the recognition result is ensured.
Compared with the prior art, the invention has the beneficial effects that:
1. the automatic and cost-saving image recognition technology can automatically process a large number of images, automatically increase the image recognizability, reduce the cost of manual intervention and the error rate, and improve the working efficiency and the recognition speed.
2. Real-time performance and accuracy, image recognition can acquire real-time data through real-time processing, compares with manual processing, promotes the rate of accuracy and the recognition effect of discernment greatly.
3. The expansibility and the intellectualization, the image recognition technology can be continuously upgraded and expanded through various algorithms and models, and the requirements of different fields can be better met.
Drawings
FIG. 1 is a block diagram of a smart identification tool based on 0 CR.
Detailed Description
In order to better understand the above technical solutions, the following detailed description will refer to the accompanying drawings and specific embodiments.
Referring to fig. 1, in the embodiment of the invention, a smart recognition tool based on 0CR includes a data acquisition module, a preprocessing module, a feature extraction and selection module, a classification decision module, and an information recognition module; the data acquisition module is used for acquiring the image file to be processed transmitted by the terminal equipment; the preprocessing module is used for preprocessing the image file to be processed, wherein the preprocessing comprises the steps of increasing the definition of the image to be processed, adjusting the angle of the image and normalizing; the feature extraction and selection module is used for extracting and selecting the same feature information of the preprocessed image file; the classification decision module is used for acquiring the information of the image to be processed and classifying the image according to the scene type; the information identification module is used for matching the classification and result of the image acquisition with the original template in the image identification information base and outputting a matching identification result in the image to be processed. The intelligent recognition tool can effectively improve recognition effect and recognition speed, and high-accuracy information recognition and classification of images are achieved through an image recognition technology.
In this embodiment: the specific process of preprocessing the image file to be processed by the preprocessing module is as follows: normalizing the size of the acquired image, and acquiring illumination intensity parameters related to the acquired image; judging whether the collected image has abnormal illumination or not based on the illumination intensity parameter, and simultaneously analyzing the abnormal type of the abnormal illumination image; carrying out illumination correction on the illumination abnormal image; performing dotting and positioning on the image file with the increased definition of the image to be processed to obtain image features with background colors; image segmentation is performed on the image based on the set of image background colors. The image file is kept available to the maximum extent by preprocessing (such as image enhancement and normalization) the input image file. The image normalization processing is a method for preprocessing image data, and aims to adjust pixel values in an image to be within a reasonable range. There are two common methods of image normalization. One is to subtract the minimum pixel value of the image from all the pixel values in the image, then divide the resulting difference by the difference between the maximum pixel value and the minimum pixel value of the image, and the result is the normalized pixel value. The other is to subtract the average pixel value of the image from all the pixel values in the image, and then divide the obtained difference by the standard deviation of the image, so that the normalized pixel value is obtained. Illumination correction of an image refers to adjusting the brightness, contrast, and hue of the image to improve the quality and readability of the image. In the process of illumination correction of an image, common algorithms include histogram equalization, adaptive histogram equalization, retinex algorithm, and the like. Pointing an image file is finding one or more points on the image and determining its location. In image processing, description point positioning is often used for image stitching, image registration, object tracking, and other tasks. By finding the feature points between the two images and determining the transformation relationship between them, the two images can be aligned and stitched into a larger image. In addition, tracking and identification of the target can be achieved by locating the position of the target in the image.
In this embodiment: the feature extraction and selection module is used for carrying out reprocessing on the image file before extracting and selecting the same feature information of the preprocessed image file, and specifically comprises the following steps: based on an image segmentation recognition result of the illumination correction image, invoking a corresponding shielding elimination algorithm to clear shielding information of image features of the object to be recognized; and calling a corresponding background splashing algorithm to perform background removing and subtracting operation on the image features of the objects to be identified, and forming feature vectors by utilizing the image features of the objects to be identified. The setting can facilitate the subsequent extraction and selection of the same characteristic information. The occlusion elimination algorithm is a commonly used algorithm in graphic rendering, and is mainly used for eliminating models occluded by other objects, so that the rendering efficiency is improved. The optimal shielding and eliminating algorithm can select visible objects to render, namely, before graphic rendering, judging which models are shielded, eliminating the shielded models, and only rendering the model which is not shielded, so that the workload of rendering is reduced. The background removing operation of the background splashing algorithm is realized mainly by calculating the projection of liquid drops on the background, and the specific steps are as follows:
calculating a projection of the droplet on the background, first, calculating a projection of the droplet on the background using hydrodynamic simulation software, which projection may represent the position, size and shape of the droplet on the background;
a mask is created, based on the projection of the drop onto the background, a mask is created for separating the drop from the background. The mask is a gray scale image in which the shape of the drop is represented as white (or high brightness) and the background is represented as black (or low brightness).
A mask is applied. A mask is applied to the input image to effect the background subtraction operation. Specifically, the mask is bit-operated from the input image, separating the drop from the background.
The color and brightness are adjusted. After the background subtraction operation, the color and brightness of the droplets and the background may need to be adjusted to make them more natural and realistic.
In this embodiment: the specific process of extracting and selecting the same characteristic information of the preprocessed image file by the characteristic extracting and selecting module is as follows: (1) Feature classification: physical features, structural features, mathematical features; (2) Feature formation generating a set of basic features from the identified image, the features so generated being called original features by calculation (when the identified object is a digital image); (3) The number of original features is probably large, a sample is represented by a low-dimensional space through a mapping (or transformation) method, the process is called feature extraction, and the mapped secondary features are linear combinations of the original features; (4) Feature selection a number of most efficient features are selected from a set of features to achieve a reduction in feature space dimension. The setting can accelerate the extraction and selection of the same characteristic information of the preprocessed image file.
In this embodiment: the specific process of classifying the images according to the scene types by the classifying decision module is as follows: preprocessing an illumination correction image, specifically: determining a difference operator, and carrying out gray level difference in the neighborhood to obtain a difference gray level map; determining a color difference valve value, and separating an image area to be identified in the differential gray scale image from background colors by using the color difference valve value; and determining a pixel value interval to which a pixel value of each background color area pixel point in the image belongs, and obtaining a first pixel identification information sequence based on a corresponding relation between the pixel value interval to which the pixel value of each pixel point belongs and preset pixel value interval identification information. The setting automatically identifies the scene type in the image and various elements in the scene by classifying and annotating different positions of the image file. The differential operator is an operator, and if Δf (x) =f (x+1) -f (x) is recorded for any real function f (x), Δ is referred to as a forward differential operator, and is simply referred to as a differential operator. The difference is one of the basic concepts of computational mathematics, and refers to the change amount of a discrete function on discrete nodes.
In this embodiment: the specific obtaining process of the first pixel identification information sequence is as follows: determining a plurality of continuous pixel value intervals divided in advance, and determining a pixel value interval to which each pixel value of each pixel point in the first area belongs; dividing the first region pixel points based on the pixel value intervals to obtain at least one first pixel point pixel value string, wherein the pixel value intervals of the pixel points in any two adjacent first pixel point pixel value strings are different; and ordering the pixel identification information of each first pixel point pixel value string to obtain a corresponding first pixel identification information sequence. The first pixel identification information sequence refers to information carried by each pixel point in the image, and is used for describing various properties of the image, such as color, brightness, contrast and the like. In image processing, a sequence of pixel identification information is typically used to perform various operations on an image, such as scaling, cropping, rotation, and the like. By processing the pixel identification information sequence, various analyses and processes of the image, such as image segmentation, feature extraction, object detection, and the like, can be realized. Therefore, the first pixel identification information sequence has an important meaning in image processing.
In this embodiment: the specific process of the information identification module for matching the classification and result of the image acquisition with the original template in the image identification information base is as follows: acquiring image features of an object to be identified, and carrying out feature extraction on the image features based on a convolution algorithm to obtain an intermediate feature matrix; sorting and reorganizing the intermediate feature matrix to obtain a post feature matrix; performing feature extraction on the post feature matrix based on a convolution algorithm to obtain an image feature set; and matching the image feature set with the original template in the image identification information base. The image recognition technique employed in the present application can customize the optimization algorithm, improve the model structure to ensure that its effect reaches a desired level, wherein the convolution algorithm is an algorithm for image processing that can apply a function (convolution kernel) to each pixel of the image and store the calculation result in a new image. This process is typically done in the frequency domain, and the computation can be accelerated by fourier transformation. In convolution algorithms, the convolution kernel is typically a small matrix whose center corresponds to each pixel in the image. Each element in the convolution kernel has a weight for calculation with the pixel value at the corresponding position in the image. This calculation is to multiply each element of the convolution kernel with the image pixel value it covers and then add the results of all the products to get the new pixel value for that location. Convolution algorithms may be used for a variety of different image processing tasks such as filtering, edge detection, feature extraction, etc. It is a very flexible and powerful tool that can be used to process a variety of different types of images and the weights of the convolution kernels can be adjusted to achieve different processing effects. The middle feature matrix refers to the feature matrix output by the convolution layer in the convolution neural network. The post feature matrix refers to the feature matrix processed by the convolution layer and the activation function in the convolution neural network.
In this embodiment: the output matching recognition result in the image to be processed is displayed through the mobile terminal and the computer terminal, and meanwhile, the manual adjustment of the recognition result is supported, so that the accuracy of the recognition result is ensured.
The invention has the advantages of automation and cost saving, and the image recognition technology can automatically process a large number of images, automatically increase the image recognizability, reduce the cost of manual intervention and error rate, and improve the working efficiency and recognition speed. In terms of instantaneity and accuracy, the image recognition of the method can acquire real-time data through real-time processing, and compared with manual processing, the method greatly improves recognition accuracy and recognition effect. In addition, the invention has expansibility and intellectualization, the image recognition technology can be continuously upgraded and expanded through various algorithms and models, and the requirements of different fields can be better met.
It will be apparent to those skilled in the art that various modifications and variations can be made to the present invention without departing from the spirit or scope of the invention. Thus, it is intended that the present invention also include such modifications and alterations insofar as they come within the scope of the appended claims or the equivalents thereof.
The foregoing description is only a preferred embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art, who is within the scope of the present invention, should make equivalent substitutions or modifications according to the technical solution of the present invention and the inventive concept thereof, and should be covered by the scope of the present invention.
Claims (8)
1. The intelligent recognition tool based on 0CR is characterized by comprising a data acquisition module, a preprocessing module, a feature extraction and selection module, a classification decision module and an information recognition module;
the data acquisition module is used for acquiring an image file to be processed transmitted by the terminal equipment; the preprocessing module is used for preprocessing the image file to be processed, wherein the preprocessing comprises the steps of increasing the definition of the image to be processed, adjusting the angle of the image and normalizing; the feature extraction and selection module is used for extracting and selecting the same feature information of the preprocessed image file; the classification decision module is used for acquiring the information of the image to be processed and classifying the image according to the scene type; the information identification module is used for matching the classification and result of the image acquisition with the original template in the image identification information base and outputting a matching identification result in the image to be processed.
2. The intelligent recognition tool based on 0CR as set forth in claim 1, wherein the preprocessing module performs preprocessing of the image file to be processed by:
normalizing the size of the acquired image, and acquiring illumination intensity parameters related to the acquired image;
judging whether the collected image has abnormal illumination or not based on the illumination intensity parameter, and simultaneously analyzing the abnormal type of the abnormal illumination image;
carrying out illumination correction on the illumination abnormal image;
performing dotting and positioning on the image file with the increased definition of the image to be processed to obtain image features with background colors;
image segmentation is performed on the image based on the set of image background colors.
3. The intelligent recognition tool based on 0CR as set forth in claim 2, wherein the feature extraction and selection module re-processes the image file before extracting and selecting the same feature information of the preprocessed image file, specifically as follows:
based on an image segmentation recognition result of the illumination correction image, invoking a corresponding shielding elimination algorithm to clear shielding information of image features of the object to be recognized;
and calling a corresponding background splashing algorithm to perform background removing and subtracting operation on the image features of the objects to be identified, and forming feature vectors by utilizing the image features of the objects to be identified.
4. A smart recognition tool based on 0CR as recited in claim 3, wherein the feature extraction and selection module extracts and selects the same feature information of the preprocessed image file by:
(1) Feature classification: physical features, structural features, mathematical features;
(2) Feature formation generating a set of basic features from the identified image, such generated features being called raw features;
(3) Extracting features, namely representing samples by a low-dimensional space through a mapping method, wherein the mapped secondary features are linear combinations of original features;
(4) Feature selection a number of most efficient features are selected from a set of features to achieve a reduction in feature space dimension.
5. The intelligent recognition tool based on 0CR as set forth in claim 4, wherein the classification decision module classifies the image according to scene type by:
preprocessing an illumination correction image, specifically: determining a difference operator, and carrying out gray level difference in the neighborhood to obtain a difference gray level map; determining a color difference valve value, and separating an image area to be identified in the differential gray scale image from background colors by using the color difference valve value;
and determining a pixel value interval to which a pixel value of each background color area pixel point in the image belongs, and obtaining a first pixel identification information sequence based on a corresponding relation between the pixel value interval to which the pixel value of each pixel point belongs and preset pixel value interval identification information.
6. The 0 CR-based intelligent recognition tool according to claim 5, wherein the specific obtaining process of the first pixel identification information sequence is:
determining a plurality of continuous pixel value intervals divided in advance, and determining a pixel value interval to which each pixel value of each pixel point in the first area belongs;
dividing the first region pixel points based on the pixel value intervals to obtain at least one first pixel point pixel value string, wherein the pixel value intervals of the pixel points in any two adjacent first pixel point pixel value strings are different;
and ordering the pixel identification information of each first pixel point pixel value string to obtain a corresponding first pixel identification information sequence.
7. The intelligent recognition tool based on 0CR as set forth in claim 6, wherein the information recognition module performs the steps of matching the classification and result of the image acquisition with the original template in the image recognition information base:
acquiring image features of an object to be identified, and carrying out feature extraction on the image features based on a convolution algorithm to obtain an intermediate feature matrix;
sorting and reorganizing the intermediate feature matrix to obtain a post feature matrix;
performing feature extraction on the post feature matrix based on a convolution algorithm to obtain an image feature set;
and matching the image feature set with the original template in the image identification information base.
8. The intelligent recognition tool based on 0CR of claim 1, wherein the output matching recognition result in the image to be processed is displayed through a mobile terminal and a computer terminal, and meanwhile, manual adjustment of the recognition result is supported to ensure the accuracy of the recognition result.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311487549.3A CN117496532A (en) | 2023-11-09 | 2023-11-09 | Intelligent recognition tool based on 0CR |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311487549.3A CN117496532A (en) | 2023-11-09 | 2023-11-09 | Intelligent recognition tool based on 0CR |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117496532A true CN117496532A (en) | 2024-02-02 |
Family
ID=89684432
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311487549.3A Pending CN117496532A (en) | 2023-11-09 | 2023-11-09 | Intelligent recognition tool based on 0CR |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117496532A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117911796A (en) * | 2024-03-18 | 2024-04-19 | 深圳市玛塔创想科技有限公司 | Intelligent data processing system and method based on image recognition |
-
2023
- 2023-11-09 CN CN202311487549.3A patent/CN117496532A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117911796A (en) * | 2024-03-18 | 2024-04-19 | 深圳市玛塔创想科技有限公司 | Intelligent data processing system and method based on image recognition |
| CN117911796B (en) * | 2024-03-18 | 2024-06-21 | 深圳市玛塔创想科技有限公司 | Intelligent data processing system and method based on image recognition |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109154978B (en) | System and method for detecting plant diseases | |
| CN111553837B (en) | Artistic text image generation method based on neural style migration | |
| CN107909081B (en) | Method for quickly acquiring and quickly calibrating image data set in deep learning | |
| CN110119741A (en) | A kind of card card image information recognition methods having powerful connections | |
| Alkoffash et al. | A survey of digital image processing techniques in character recognition | |
| CN109711268B (en) | Face image screening method and device | |
| CN111680690A (en) | Character recognition method and device | |
| CN113435407B (en) | Small target identification method and device for power transmission system | |
| CN117496532A (en) | Intelligent recognition tool based on 0CR | |
| CN113537037A (en) | Pavement disease identification method, system, electronic device and storage medium | |
| CN112417955A (en) | Patrol video stream processing method and device | |
| CN113052215A (en) | Sonar image automatic target identification method based on neural network visualization | |
| CN117557784B (en) | Target detection method, target detection device, electronic equipment and storage medium | |
| CN112348018A (en) | Digital display type instrument reading identification method based on inspection robot | |
| CN116843581B (en) | Image enhancement method, system, device and storage medium for multi-scene graph | |
| CN113989814A (en) | Image generation method and device, computer equipment and storage medium | |
| CN110188693B (en) | Improved complex environment vehicle feature extraction and parking discrimination method | |
| CN113435358B (en) | Sample generation method, device, equipment and program product for training model | |
| Krishna et al. | Digital image processing techniques in character recognition-a survey | |
| CN115187788A (en) | Crop seed automatic counting method based on machine vision | |
| CN115063679A (en) | Pavement quality assessment method based on deep learning | |
| CN110276260B (en) | Commodity detection method based on depth camera | |
| Rani et al. | Object Detection in Natural Scene Images Using Thresholding Techniques | |
| CN113012167A (en) | Combined segmentation method for cell nucleus and cytoplasm | |
| CN111242047A (en) | Image processing method and apparatus, electronic device, and computer-readable 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 |