Artificial intelligence machine vision image acquisition device

Abstract

The invention discloses an artificial intelligence machine vision image acquisition device, and relates to the technical field of image acquisition. The artificial intelligence machine vision image acquisition device comprises an image acquisition unit, an image processing unit, an image optimization unit, a brightness compensation unit, a motion control unit and a control unit, the brightness compensation unit is arranged to firstly monitor acquired ambient light, and according to acquired brightness parameters, the image optimization unit is arranged to optimize the image according to the acquired brightness parameters. Obtaining an optimal light source output voltage parameter subset matched with the corresponding ambient brightness in the training parameter set database, obtaining an optimal light source output voltage parameter in the optimal light source output voltage parameter subset, and sending a control instruction to a brightness regulator through a control unit, the brightness regulator adjusts the light emitting brightness of the light supplementing lamp by changing the voltage of the light supplementing lamp, and the optimal light supplementing degree is matched, so that the illumination light of the image acquisition target area is uniform, the image acquisition target area is uniformly exposed, and the quality of the acquired image is ensured.

Description

Technical Field

[0001] The present invention relates to the field of image acquisition technology, and in particular to an artificial intelligence machine vision image acquisition device. Background Art

[0002] At present, image acquisition mostly relies on manual camera acquisition, which consumes a lot of manpower and has low work efficiency. The current intelligent visual image acquisition device has a simple structure, cannot collect image information at a regular time, cannot capture the position of the image, so the collected image information is not complete and clear, cannot realize the self-start of the device, and cannot realize remote control, which is very inconvenient to use.

[0003] The invention with publication number CN104980658A discloses an intelligent visual image acquisition device, comprising a processor, wherein the input end of the processor is electrically connected to a battery, a timer, a laser locator and an output end of an image processing module, the input end of the image processing module is electrically connected to the output end of an information acquisition module, the output end of the processor is electrically connected to the input end of a display, and the wireless communication module is bidirectionally electrically connected to a remote control end.

[0004] As shown in the above invention, the existing image acquisition device is equipped with a laser locator, which can locate the picture to be collected, so that the picture collection is more comprehensive and clearer. However, it is very difficult to collect pictures in complex environments. For example, it is very difficult to collect images in scenes full of obstructions, dim light, uneven lighting and exposure, and it is impossible to obtain accurate image information. Summary of the invention

[0005] In view of the deficiencies in the prior art, the present invention provides an artificial intelligence machine vision image acquisition device to solve the existing problems.

[0006] To achieve the above objectives, the present invention is implemented through the following technical solutions: an artificial intelligence machine vision image acquisition device, comprising:

[0007] An image acquisition unit, used for acquiring image data of a target;

[0008] An image processing unit, the image processing unit is electrically connected to the image acquisition unit and is used to perform preliminary processing on the acquired image;

[0009] An image optimization unit, the image optimization unit is electrically connected to the image processing unit and is used to further optimize the image after preliminary processing;

[0010] The brightness compensation unit is mainly composed of a fill light, a brightness adjuster and a brightness sensor, and is used to adjust the brightness of the fill light to perform illumination compensation on the image;

[0011] The motion control unit is mainly composed of a horizontal sliding mechanism, a vertical sliding mechanism and a distance sensor, and is used to adjust the position of the industrial camera on the image acquisition unit so that the industrial camera can obtain image data of the target from different positions and directions;

[0012] A control unit is electrically connected to the image acquisition unit, the brightness compensation unit and the motion control unit, and is used to control image acquisition, brightness control and position adjustment.

[0013] Preferably, the acquisition device also includes a big data module for acquiring light source output voltage, historical ambient brightness data, learning and training, and establishing a training parameter set database based on big data analysis.

[0014] Preferably, the image acquisition unit is mainly composed of an industrial camera and a ranging sensor, and the industrial camera and the ranging sensor cooperate to achieve automatic zoom when acquiring target image data.

[0015] Preferably, the image processing unit is mainly composed of an image enhancement module, an image sharpening module and an image smoothing module, and the image enhancement module, the image edge sharpening processing module and the image smoothing processing module are connected in sequence. The image enhancement module is used to adjust the contrast of the image, the image edge sharpening processing module is used to enhance the contour edges and details in the image to form a complete object boundary, and the image smoothing processing module is used to make the image brightness change smoothly, reduce the sudden gradient, and improve the image quality.

[0016] Preferably, the image optimization unit is mainly composed of an image segmentation module, a target acquisition module and an image analysis and optimization module, and the image segmentation module, the target acquisition module and the image analysis and optimization module are connected in sequence.

[0017] Preferably, the image segmentation module is used to segment the image to obtain a background image and a target image, the target acquisition module is used to obtain the target image, and the image analysis and optimization module is used to analyze the target image, obtain the inclusion rate of the target image based on the analysis results, and generate an optimized target image based on the inclusion rate.

[0018] Preferably, the brightness adjuster adopts a thyristor dimmer, and the brightness sensor adopts a LI-210R visible light illumination sensor.

[0019] Preferably, the control unit adopts a STM32F103C8T6 control chip.

[0020] Beneficial Effects

[0021] The present invention provides an artificial intelligence machine vision image acquisition device. Compared with the prior art, it has the following beneficial effects:

[0022] 1. The artificial intelligence machine vision image acquisition device is provided with a brightness compensation unit to first monitor the ambient light for acquisition, and obtains the best light source output voltage parameter subset that matches the corresponding ambient brightness in the training parameter set database according to the acquired brightness parameters, obtains the best light source output voltage parameter in the best light source output voltage parameter subset, and sends a control instruction to the brightness adjuster through the control unit. The brightness adjuster adjusts the luminous brightness of the fill light by changing the voltage of the fill light to match the best fill light, thereby ensuring uniform illumination of the image acquisition target area, uniformly exposing the image acquisition target area, and ensuring the quality of the acquired image.

[0023] 2. The artificial intelligence machine vision image acquisition device, by setting an image processing unit and an image optimization unit, the image processing unit pre-processes the acquired image, the image optimization unit segments the image to obtain a background image and a target image, sends the target image to a target image acquisition module to obtain the target image, segments the image multiple times to obtain the target image, better processes the unclear image multiple times to obtain a clear image, and then analyzes the target image, obtains the inclusion rate of the target image based on the analysis result, and generates an optimized target image based on the inclusion rate, thereby further improving the quality of the target image. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a schematic diagram of system connection of the present invention;

[0025] Figure 2 It is a flow chart of the visual image acquisition method of the present invention;

[0026] Figure 3 It is a schematic diagram of the flow chart of the visual image processing method of the present invention. DETAILED DESCRIPTION

[0027] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0028] See Figure 1-3 , the present invention provides the following four technical solutions:

[0029] The first embodiment: an artificial intelligence machine vision image acquisition device, comprising:

[0030] An image acquisition unit is used to acquire image data of a target. The image acquisition unit is mainly composed of an industrial camera and a ranging sensor. The industrial camera and the ranging sensor cooperate to realize automatic zooming when acquiring target image data.

[0031] An image processing unit, the image processing unit is electrically connected to the image acquisition unit and is used to perform preliminary processing on the acquired image. The image processing unit is mainly composed of an image enhancement module, an image sharpening module and an image smoothing module. The image enhancement module, the image edge sharpening processing module and the image smoothing processing module are connected in sequence. The image enhancement module is used to adjust the contrast of the image. The image edge sharpening processing module is used to enhance the contour edges and details in the image to form a complete object boundary. The image smoothing processing module is used to make the image brightness change smoothly and gradually, reduce the sudden gradient, and improve the image quality.

[0032] An image optimization unit, the image optimization unit is electrically connected to the image processing unit, and is used to further optimize the image after preliminary processing. The image optimization unit is mainly composed of an image segmentation module, a target acquisition module, and an image analysis and optimization module. The image segmentation module, the target acquisition module, and the image analysis and optimization module are connected in sequence. The image segmentation module is used to segment the image to obtain a background image and a target image. The target acquisition module is used to obtain a target image. The image analysis and optimization module is used to analyze the target image, obtain the inclusion rate of the target image according to the analysis result, and generate an optimized target image according to the inclusion rate;

[0033] The brightness compensation unit is mainly composed of a fill light, a brightness adjuster and a brightness sensor, and is used to adjust the brightness of the fill light to perform illumination compensation on the image. The brightness adjuster adopts a thyristor dimmer, and the brightness sensor adopts a LI-210R visible light illumination sensor;

[0034] The motion control unit is mainly composed of a horizontal sliding mechanism, a vertical sliding mechanism and a distance sensor, and is used to adjust the position of the industrial camera on the image acquisition unit so that the industrial camera can obtain image data of the target from different positions and directions;

[0035] A control unit, wherein the control unit is electrically connected to the image acquisition unit, the brightness compensation unit and the motion control unit, and is used to control image acquisition, brightness control and position adjustment. The control unit adopts an STM32F103C8T6 control chip;

[0036] It also includes a big data module for acquiring light source output voltage, historical ambient brightness data, learning and training, and establishing a training parameter set database based on big data analysis.

[0037] Second implementation mode: Another embodiment of the present invention provides an artificial intelligence machine vision image acquisition method, which includes the following steps:

[0038] Step 1: The image data of the target is obtained through the industrial camera on the image acquisition unit. The ranging sensor cooperates with the industrial camera to achieve automatic zoom;

[0039] Step 2: Calculate the brightness of the target image, obtain the light ratio of the target image according to the brightness, and then determine whether the acquisition scene of the target image is a large light ratio scene

[0040] Step 3: When the acquisition scene is a high-light ratio scene, the accumulation degree of each pixel value is obtained according to the number of each pixel value in the dark area of ​​the grayscale histogram of the target image, and the accumulation pixel points whose accumulation degree is greater than the accumulation threshold are screened out;

[0041] Step 4: Obtain the changed pixels with hierarchical changes in the connected domain by calculating the pixel value changes between each accumulated pixel point and its neighboring pixel points, and obtain the acquired image according to the changed pixels and the parameters of the acquisition device.

[0042] The judgment process of the high light ratio scene is as follows:

[0043] A light ratio threshold is set. When the light ratio is greater than the light ratio threshold, the captured scene is a large light ratio scene.

[0044] The step of acquiring the changed pixels comprises:

[0045] When the difference between the accumulated pixel point and any of its neighboring pixels is greater than a difference threshold, a change ratio of the neighboring pixels of the accumulated pixel point is obtained. When the change ratio is greater than the change threshold, the accumulated pixel point is a changed pixel.

[0046] A third embodiment of the present invention provides an artificial intelligence machine vision image processing method, which includes the following steps:

[0047] S1: Preliminary processing of the obtained image is performed through an image enhancement module, an image sharpening module and an image smoothing module;

[0048] S2: segmenting the preliminarily processed image through an image segmentation module to generate a small-block target image database and a small-block background image database, and automatically integrating the small-block target images in the target image database to generate a target image;

[0049] S3: The generated target image is acquired through the target acquisition module and sent to the target image analysis and optimization module for analysis.

[0050] The segmentation method of the image segmentation module is as follows: the image is divided into N small square images with an area of ​​n, where n is a positive number, the N small square images with an area of ​​n are processed respectively, the grayscale values ​​ω of the N small square images are obtained respectively, and the images are registered according to the grayscale values, the N small square images are converted into N small square grayscale images, the grayscale value threshold is set to 0≤§≤255, the grayscale values ​​of the N small square grayscale images are compared with the grayscale value threshold, if the grayscale value of the small square grayscale image is 0≤ω≤§, then the small square grayscale image is stored in the sub-database, if the grayscale value of the small square grayscale image is §<ω≤255, then the small square grayscale image is stored in the database, and a database of the target image is generated;

[0051] The small square grayscale images in the sub-database are integrated to generate a background image, and the small square grayscale images in the database are integrated to generate a target image.

[0052] The analysis process of the image analysis optimization module includes: storing the reference image in the cloud in advance, obtaining the inclusion rate of the target image according to the reference image, marking the reference image as U and the target image as V, establishing a rectangular coordinate system for the target image, and intercepting it, intercepting it as the initial x position, the initial y position and the interception length, obtaining the intercepted target image after the interception is completed, displaying the intercepted target image, and obtaining the entropy value of the intercepted target image;

[0053] Traverse each row and column of the reference image, intercept a movable reference image of the same size as the intercepted target image, record it as a moving reference image, and obtain the entropy value of the moving reference image. According to the entropy value of the intercepted target image and the entropy value of the moving reference image, obtain the probability density function corresponding to the reference image and the target image, and mark them as f(V) and f(U) respectively;

[0054] According to the probability density function, the joint probability density function of the reference image and the target image is obtained and marked as ρ(V,U);

[0055] According to the probability density function and joint probability density function corresponding to the reference image and the target image, the inclusion rate ρ(V,U) of the reference image and the target image is obtained; that is, the formula is:

[0056]

[0057] Set an inclusion rate threshold, compare the obtained inclusion rate with the inclusion rate threshold, and determine whether the target image needs further optimization. If the inclusion rate is less than or equal to the inclusion rate threshold, further optimization is required, and its information generation signal is sent to the management terminal to receive the target image and mark it as the optimized target image. If the inclusion rate is greater than the inclusion rate threshold, no further optimization is required.

[0058] The optimization process of the image analysis optimization module includes: evaluating the percentage of the inclusion rate of the optimization target image that is lower than the inclusion rate threshold, and obtaining an evaluation probability value;

[0059] Setting an evaluation probability value threshold, evaluating the percentage of the inclusion rate of the optimization target image that is lower than the inclusion rate threshold, and obtaining an evaluation probability value;

[0060] If the evaluation probability value of the optimized target image is greater than or equal to the evaluation probability value threshold, the optimized target image is sent back to the image processing module for further processing, and is processed by the image processing module multiple times until the inclusion rate of the obtained target image is lower than the inclusion rate threshold;

[0061] If the evaluation probability value of the optimized target image is lower than the evaluation probability value threshold, manual optimization of the target image is required.

[0062] Fourth implementation mode: Another embodiment of the present invention provides an artificial intelligence machine vision image acquisition platform, including a memory, a processor, and a computer program stored in the memory and executable on the processor.

[0063] Meanwhile, the contents not described in detail in this specification belong to the prior art known to those skilled in the art, and the model parameters of each electrical appliance are not specifically limited, and conventional equipment can be used.

[0064] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device.

[0065] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An artificial intelligence machine vision image acquisition device, characterized in that: include: An image acquisition unit, used for acquiring image data of a target; An image processing unit, the image processing unit is electrically connected to the image acquisition unit and is used to perform preliminary processing on the acquired image; An image optimization unit, the image optimization unit is electrically connected to the image processing unit and is used to further optimize the image after preliminary processing; The brightness compensation unit is mainly composed of a fill light, a brightness adjuster and a brightness sensor, and is used to adjust the brightness of the fill light to perform illumination compensation on the image; The motion control unit is mainly composed of a horizontal sliding mechanism, a vertical sliding mechanism and a distance sensor, and is used to adjust the position of the industrial camera on the image acquisition unit so that the industrial camera can obtain image data of the target from different positions and directions; A control unit is electrically connected to the image acquisition unit, the brightness compensation unit and the motion control unit, and is used to control image acquisition, brightness control and position adjustment.

2. The artificial intelligence machine vision image acquisition device according to claim 1, characterized in that: The acquisition device also includes a big data module, which is used to obtain the output voltage of the light source and the historical ambient brightness data, learn and train, and establish a training parameter set database based on big data analysis.

3. The artificial intelligence machine vision image acquisition device according to claim 1, characterized in that: The image acquisition unit is mainly composed of an industrial camera and a distance measuring sensor. The industrial camera and the distance measuring sensor cooperate to achieve automatic zoom when acquiring target image data.

4. The artificial intelligence machine vision image acquisition device according to claim 1, characterized in that: The image processing unit is mainly composed of an image enhancement module, an image sharpening module and an image smoothing module. The image enhancement module, the image edge sharpening processing module and the image smoothing processing module are connected in sequence. The image enhancement module is used to adjust the contrast of the image. The image edge sharpening processing module is used to enhance the contour edges and details in the image to form a complete object boundary. The image smoothing processing module is used to make the image brightness change smoothly, reduce the sudden gradient, and improve the image quality.

5. The artificial intelligence machine vision image acquisition device according to claim 1, characterized in that: The image optimization unit is mainly composed of an image segmentation module, a target acquisition module and an image analysis and optimization module, and the image segmentation module, the target acquisition module and the image analysis and optimization module are connected in sequence.

6. The artificial intelligence machine vision image acquisition device according to claim 5, characterized in that: The image segmentation module is used to segment the image to obtain a background image and a target image. The target acquisition module is used to obtain a target image. The image analysis and optimization module is used to analyze the target image, obtain the inclusion rate of the target image based on the analysis result, and generate an optimized target image based on the inclusion rate.

7. The artificial intelligence machine vision image acquisition device according to claim 1, characterized in that: The brightness adjuster adopts a thyristor dimmer, and the brightness sensor adopts a LI-210R visible light illumination sensor.

8. The artificial intelligence machine vision image acquisition device according to claim 1, characterized in that: The control unit adopts the STM32F103C8T6 control chip.

Images