CN102759530A - Online detection device for surface quality image - Google Patents
Online detection device for surface quality image Download PDFInfo
- Publication number
- CN102759530A CN102759530A CN2012102277306A CN201210227730A CN102759530A CN 102759530 A CN102759530 A CN 102759530A CN 2012102277306 A CN2012102277306 A CN 2012102277306A CN 201210227730 A CN201210227730 A CN 201210227730A CN 102759530 A CN102759530 A CN 102759530A
- Authority
- CN
- China
- Prior art keywords
- image
- eigenwert
- surface quality
- image block
- block
- 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.)
- Granted
Links
Images
Landscapes
- Image Processing (AREA)
Abstract
The invention provides an online detection device for a surface quality image, comprising an image collector and an image processing IP (Internet Protocol) core, wherein the image collector is used for collecting the surface quality image and transmitting the surface quality image to the image processing IP core; and the image processing IP core is used for dividing the surface quality image into a plurality of image blocks, calculating a characteristic value of each image block in parallel and judging whether the detection image block is qualified or not according to the characteristic value. The image processing IP core disclosed by the invention can divide the collected surface quality image into a plurality of the image blocks and can calculate the characteristic values of the image blocks in parallel, therefore, the online detection device has good instantaneity.
Description
Technical field
The present invention relates to technical field of image processing, relate to a kind of surface quality image on-line measuring device in particular.
Background technology
At present; Machine vision technique is widely used for fields such as quality control; Machine vision technique is followed the development of computer technology, field bus technique, and technology is ripe day by day, the promotion that the high speed development of machine vision technique is strong the development and the utilization of surface quality online measuring technique.For example, in iron and steel tandem rolling and casting process, the billet surface quality testing is one of gordian technique of robotization steel rolling cast steel, and the Flame Image Process in the quality testing of " online " billet surface is wherein emphasis and difficult point.
In the online detection of surface quality, the traditional image disposal route is realized by algorithm software, and is present; The data output rate of most high speed linear array camera exceeds Gbps; When the traditional image disposal route is exported in the high data of processing, can only rely on the dominant frequency that improves computer CPU to improve the speed that view data is handled, and the dominant frequency of computer is owing to the restriction that receives hardware; Be very limited, so the traditional image disposal route can not satisfy the real-time requirement of Flame Image Process and Algorithm Analysis.
Summary of the invention
In view of this, in order to solve the problem that the traditional image disposal route can not satisfy the real-time requirement of Flame Image Process and Algorithm Analysis, a kind of surface quality on-line measuring device is provided, technical scheme is following:
The device of the online detection of a kind of surface quality image comprises image acquisition device and image processing IP core;
Said image acquisition device is used for the collection surface quality image, and said surface quality image is transferred to said image processing IP core;
Said image processing IP core is used for said surface quality image is divided into a plurality of image blocks, and parallel computation goes out the eigenwert of each said image block, and whether detect said image block according to said eigenwert qualified.
Preferably, in the device of the online detection of above-mentioned surface quality image, said image processing IP core comprises image block pre-service and sheet menu unit, and the image buffer storage processing unit;
Said image block pre-service and sheet menu unit are used for the said surface quality image of pre-service, and said surface quality image are divided into a plurality of image blocks, transfer to said image buffer storage processing unit;
Whether said image buffer storage processing unit be used for the eigenwert of each said image block of parallel computation, and it is qualified to detect said image according to said eigenwert.
Preferably, in the device of the online detection of above-mentioned surface quality image, said image pre-service and sheet menu unit comprise the chip selection signal generator, and the image pretreatment unit;
Said image pretreatment unit is used to realize that the pretreatment operation of said surface quality image, said pretreatment operation comprise the taking advantage of, add of view data of said surface quality image at least, and delay operation;
Said chip selection signal generator; Be used to produce chip selection signal; Said chip selection signal is for being divided into the image block of n/k piece with a two field picture along the direction of motion of said surface quality image, and said image block is transferred to the signal of said image buffer storage processing unit, wherein; N is the pixel of said image acquisition device, k behavior one two field picture of said surface quality image.
Preferably, in the device of the online detection of above-mentioned surface quality image, said image buffer storage processing unit comprises image buffer storage unit and graphics processing unit;
Said image buffer storage unit is used for the said image block of buffer memory, and said image block is offered said graphics processing unit;
Said graphics processing unit; Be used to realize the eigenwert of the said image block of parallel computation; And the threshold interval of the normal surface image feature value of said eigenwert and prevision compared, if said eigenwert is not in the said threshold interval, then be judged to be the fault image block.
Preferably, in the device of the online detection of above-mentioned surface quality image, said image processing IP core also comprises the Image Sharing storage unit;
Said Image Sharing storage unit is used to store the image block of the eigenwert and the said fault of said image block, produces simultaneously to be used for the image block of the eigenwert of said image block and said the fault full up or half-full look-at-me to said industrial computer transmission.
Preferably, in the device of the online detection of above-mentioned surface quality image, said Image Sharing storage unit comprises: eigenwert interruption generating device, eigenwert storage unit, defect image storage unit, and defect image interruption generating device;
Said eigenwert storage unit is used to store all kinds of eigenwerts that said image buffer storage processing unit calculates;
When the half of the said eigenwert of said eigenwert cell stores to himself capacity, said eigenwert interruption generating device produces the first half-full look-at-me;
When the said eigenwert of institute's eigenwert cell stores during to himself capacity whole, said eigenwert interruption generating device produces the first full up look-at-me;
Said defect image storage unit is used to store the image block of said fault;
When the data of the image block of the said fault of the said defect image cell stores half to himself capacity, said defect image interruption generating device produces the second half-full look-at-me;
When the data of the image block of the said fault of said defect image cell stores during to himself capacity whole, said defect image interruption generating device produces the second full up look-at-me.
Preferably, in the device of the online detection of above-mentioned surface quality image, said device also comprises the image measurement unit, and said image measurement unit comprises image function test cell and image property test cell;
Said image function test cell, whether the eigenwert that is used to detect said image block is calculated correctly;
Said image property test cell, whether the computing velocity of eigenwert that is used to detect said image block is more than or equal to the picking rate of said surface quality image.
Have following beneficial effect in the technique scheme:
Image processing IP core provided by the invention can be divided into a plurality of image blocks, the eigenwert of parallel computation image block with collecting the surface quality image; Be a real-time clock, through test, the monolithic data processed result of image processing IP core provided by the invention comes out only to lag behind more than 20 clock period of image input; If the major clock of image processing IP core is 250MHz; Also be 20* (1/250MHz)=80ns, for the two field picture processing time 80ns*128=10.24us of (generally being divided into 128), and 10.24us is much smaller than the acquisition time of a two field picture; Be system quasi real time theoretically; From practical applications is strict real-time system, has good real time performance, and CPU that this is computer etc. is incomparable.
Description of drawings
In order to be illustrated more clearly in the embodiment of the invention or technical scheme of the prior art; To do to introduce simply to the accompanying drawing of required use in embodiment or the description of the Prior Art below; Obviously, the accompanying drawing in describing below only is embodiments of the invention, for those of ordinary skills; Under the prerequisite of not paying creative work, can also obtain other accompanying drawing according to the accompanying drawing that provides.
A kind of structural representation of the surface quality image on-line measuring device that Fig. 1 provides for the embodiment of the invention;
Another structural representation of the surface quality image on-line measuring device that Fig. 2 provides for the embodiment of the invention;
Another structural representation of the surface quality image on-line measuring device that Fig. 3 provides for the embodiment of the invention;
Another structural representation of the surface quality image on-line measuring device that Fig. 4 provides for the embodiment of the invention;
The structural framing figure of the image detecting element that Fig. 5 provides for the embodiment of the invention;
The surface quality image on-line measuring device that Fig. 6 provides for the embodiment of the invention carries out the synoptic diagram of piecemeal to image;
Surface quality image on-line measuring device operation time sequence figure in FPGA that Fig. 7 provides for the embodiment of the invention.
Embodiment
To combine the accompanying drawing in the embodiment of the invention below, the technical scheme in the embodiment of the invention is carried out clear, intactly description, obviously, the embodiment that describes that gathers only is the present invention's part embodiment, rather than whole embodiment.Based on the embodiment among the present invention, those of ordinary skills are not making the every other embodiment that is obtained under the creative work prerequisite, all belong to the scope of the present invention's protection.
With reference to figure 1, a kind of device that the online detection of surface quality is provided of Fig. 1 signal comprises: image acquisition device U1 and image processing IP core U2.
Image acquisition device U1 is used for the collection surface quality image, and above-mentioned surface quality image is transferred to above-mentioned image processing IP core;
Image processing IP core U2 is used for above-mentioned surface quality image is divided into a plurality of image blocks, and parallel computation goes out the eigenwert of each above-mentioned image block, and whether detect above-mentioned image block according to above-mentioned eigenwert qualified.
Line-scan digital camera is the common image capturing device of surperficial detection range, is used for the imaging and the analysis of high-speed moving object.Line-scan digital camera can be used as image acquisition device collection surface quality image, and the surface quality image that collects is transferred to image processing IP core U2 handles.
Image processing IP core provided by the invention can be divided into a plurality of image blocks, the eigenwert of parallel computation image block with collecting the surface quality image; Be a real-time system, through test, the monolithic data processed result of image processing IP core provided by the invention comes out only to lag behind more than 20 clock period of image input; If the major clock of image processing IP core is 250MHz; Also be 80ns, for the two field picture processing time 80ns*128=10.24us of (generally being divided into 128), and 10.24us is much smaller than the acquisition time of a two field picture; Being system quasi real time theoretically, is strict real-time system from practical applications.With line-scan digital camera line frequency 20KHz is example; According to method in the embodiment, the image frame grabber time (times of 32 row)=32* (1/20KHz)=1.6ms, it is far longer than 10.24us; Have good real time performance, CPU or DSP that this is computer etc. is incomparable.
With reference to figure 2, a kind of structural representation that image processing IP core U2 is provided of Fig. 2 signal.
Image processing IP core U2 can comprise the U21 of image block pre-service and sheet menu unit, image buffer storage processing unit U22.
Image block pre-service and the U21 of sheet menu unit are used for the above-mentioned surface quality image of pre-service, and above-mentioned surface quality image are divided into a plurality of image blocks, transfer to above-mentioned image buffer storage processing unit; Whether image buffer storage processing unit U22 be used for the eigenwert of each above-mentioned image block of parallel computation, and it is qualified to detect above-mentioned image according to above-mentioned eigenwert.
Image block pre-service and the U21 of sheet menu unit can comprise chip selection signal generator U211, and image pretreatment unit U212;
Image pretreatment unit U212 is used to realize that the pretreatment operation of above-mentioned surface quality image, above-mentioned pretreatment operation comprise the taking advantage of, add of view data of above-mentioned surface quality image at least, and delay operation;
Chip selection signal generator U211; Be used to produce chip selection signal; Above-mentioned chip selection signal is for being divided into the image block of n/k piece with a two field picture along the direction of motion of above-mentioned surface quality image, and above-mentioned image block is transferred to the signal of above-mentioned image buffer storage processing unit, wherein; N is the pixel of above-mentioned image acquisition device, k behavior one two field picture of above-mentioned surface quality image.
Image buffer storage processing unit U22 comprises image buffer storage unit U221 and graphics processing unit U222; Image buffer storage unit U221 is used for the above-mentioned image block of buffer memory, and above-mentioned image block is offered above-mentioned graphics processing unit; Graphics processing unit U222; Be used to realize the eigenwert of the above-mentioned image block of parallel computation; And the threshold interval of the normal surface image feature value of above-mentioned eigenwert and prevision compared, if above-mentioned eigenwert is not in the above-mentioned threshold interval, then be judged to be the fault image block.Chip selection signal generator U211 produces chip selection signal, and the image block that chip selection signal is corresponding with a two field picture transfers to image buffer storage unit U221, and wherein, the number of image buffer storage unit U221 and graphics processing unit U222 is n/k, also is the number of image block.
The image block of graphics processing unit U222 failure judgement can may further comprise the steps:
To calculate to such an extent that above-mentioned image block eigenwert and normal surface image feature value interval compare,, then be judged to be the fault image block if be not in the above-mentioned threshold interval.
Be that object describes below with the billet surface, can choose variance or light intensity average is eigenwert.
Choosing variance is eigenwert:
Method: normal billet surface variance threshold values interval is≤4 through repeatedly training the back; The all images piece can be sent into the image buffer storage processing unit so; Calculate variance yields through graphics processing unit>4 image block, then be judged to be the fault image block, all the other are normal billet surface.
Choosing the light intensity average is eigenwert:
Method: the normal strong average threshold interval of steel billet surface light is through after repeatedly training being 35≤normal light intensity≤100; The all images piece can be sent into the image buffer storage processing unit so; Calculate the image block of light intensity average < 35 perhaps>100 through graphics processing unit; Then be judged to be the fault image block, all the other are normal billet surface.
Need to prove, because a category feature value is very sensitive to some defective, but to part defective and insensitive; So whether remove to define imaging surface qualified through single eigenwert often is not very reliable; Therefore, in the process of practical application, can calculate the various features value; Get the common factor of a plurality of eigenwert threshold intervals, then can bring up to the fault recall rate more than 95%!
With reference to figure 3, image processing IP core U2 can also comprise Image Sharing storage unit U23.
Image Sharing storage unit U23 is used to store the image block of the eigenwert and the above-mentioned fault of above-mentioned image block, produces simultaneously to be used for the image block of the eigenwert of above-mentioned image block and above-mentioned the fault full up or half-full look-at-me to above-mentioned industrial computer transmission.
Image Sharing storage unit U23 can comprise: eigenwert interruption generating device U231, eigenwert storage unit U232, defect image storage unit U233, and defect image interruption generating device U234.
Graphics processing unit U222 among the image buffer storage processing unit U22 is stored in eigenwert and defect image piece respectively among eigenwert storage unit U232 and the defect image interruption generating device U234.
Eigenwert interruption generating device U231 is used to store all kinds of eigenwerts that above-mentioned image buffer storage processing unit calculates; When eigenwert storage unit U232 stores the half of above-mentioned eigenwert to himself capacity, eigenwert interruption generating device U231 produces the first half-full look-at-me; When the above-mentioned eigenwert of eigenwert storage unit U232 during to himself capacity whole, eigenwert interruption generating device U231 produces the first full up look-at-me; The first half-full look-at-me and the first full up look-at-me supply the main equipment of image processing IP core to start the data that read eigenwert.Eigenwert interruption generating device can cooperate main equipment to accomplish the ping-pong operation of data transmission easily.
Defect image storage unit U233 is used to store the image block of above-mentioned fault; The data of image block of storing above-mentioned fault as defect image storage unit U233 are to a half of himself capacity, and above-mentioned defect image interruption generating device U234 produces the second half-full look-at-me; The data of image block of storing above-mentioned fault as defect image storage unit U233 are during to himself capacity whole, and above-mentioned defect image interruption generating device U234 produces the second full up look-at-me; The second half-full look-at-me and the second full up look-at-me supply the main equipment of image processing IP core to start and read the defect image data.Defect image interruption generating device can cooperate main equipment to accomplish the ping-pong operation of data transmission easily.Ping-pong operation is the common a kind of operation of electronic applications, and method is that two storage areas act on simultaneously, and storage and transmission are alternately carried out; B piece transmission when being the storage of A piece; The transmission of A piece reaches the purpose of trading space for time in the time of the storage of B piece, is very suitable for the design of real-time system.
Design feature value interruption generating device and defect image interruption generating device in Image Sharing storage unit U23; Eigenwert interruption generating device produces half-full signal of eigenwert or the full up signal of eigenwert; Defect image interruption generating device produces the half-full signal of defect image; In order to accomplish the ping-pong operation of data transmission, in order to promote real-time.
Eigenwert storage unit U232 and defect image storage unit U233 all provide PLB (Processor Local Bus) or OPB (On-chip Peripheral Bus) bus access interface; Eigenwert storage unit U232 and defect image storage unit U233 all can through PLB or OPB EBI easily carry on main equipment, accomplish the transmission of data to industrial computer.The main equipment here can be for being embedded into FPGA inner soft-core processor MicroBlaze or stone processor P owerPC.Above-mentioned PLB or OPB EBI are the open main equipments of giving, and main equipment can come processing transactions according to look-at-me, give industrial computer such as starting PCIe transmission feature value or defective data.The PCIe transmission here is through dma mode, and the data rate test has reached 1.48Gbps at present.
Further, the present invention also comprises image measurement unit U3, and image measurement unit U3 comprises image function test cell U31 and image property test cell U32.
Image function test cell U31, whether the eigenwert that is used to detect above-mentioned image block is calculated correctly;
Image property test cell U32, whether the computing velocity of eigenwert that is used to detect above-mentioned image block is more than or equal to the picking rate of above-mentioned surface quality image.
Image measurement unit U3 is used for handling at the preliminary stage test pattern function of IP kernel U2, and is concrete, detect the eigenwert of above-mentioned image block and whether calculate correctly, eigenwert comprise image block such as characteristics such as variance, entropy, gray scale, texture and frequency spectrums.Image measurement unit U3 test pattern is handled the performance of IP kernel U2, is specially, and whether the computing velocity of eigenwert that detects above-mentioned image block is more than or equal to the picking rate of above-mentioned surface quality image.The surface quality image to be tested that U3 input in image measurement unit collects; The eigenwert of these images is that the technician knows in advance; The eigenwert of eigenwert of calculating from image processing IP core U2 and surface quality image to be tested is compared, and is consistent then explain that the function of this image processing IP core is correct; If the computing velocity of the eigenwert of above-mentioned image block is more than or equal to the picking rate of above-mentioned surface quality image; That is to say; If before can getting into image processing IP core at the next frame of surface image data; Accomplish the eigenvalue calculation of surface image data present frame, and with the eigenwert of present frame and image data transmission to industrial computer, explain that then the performance of this image processing IP core meets the demands.The structural framing of image measurement unit U3 is as shown in Figure 5, can accomplish the function and the performance test of above-mentioned image processing IP core.
Wherein, 301 is the reset signal input of image measurement unit; 302 is the enable signal input of image measurement unit; 303 is logic sum gate; 304 and 312 is logical AND gate; 305 and 310 is totalizer; 307 is first comparer, and comparison expression is a>b; 306 is real constant 4096; 308 is logic inverter; 309 is real constant 4094; 311 is second comparer, and comparison expression is a <b; 313 is the test data of image measurement unit in the Matlab platform; 314 is that view data imports the inner input port of FPGA from the Matlab platform; 315 are reset signal output; 316 are the output of row useful signal; 317 are view data output.
Wherein, the t in above-mentioned 313 is that valid data are capable, aligns with 316 high level " 1 "; Y is an invalid data, is set to 0 here, aligns with 316 low level " 0 ".
Image measurement unit U3 is in order to obtain the signal consistent with the line-scan digital camera output timing: LINE_VALID and DATA_OUT.The RST_OUT signal is in order to debug usefulness, let line-scan digital camera can with the image processing IP core synchronous reset.Image measurement unit U3 has following effect, first: the function and the performance of can test pattern gathering IP kernel after the function and the performance of IMAQ IP kernel satisfy condition, really during images acquired, are replaced by line-scan digital camera; The second,, project is done early stage assessment and the progress propelling has very large meaning because line-scan digital camera costs an arm and a leg and procurement cycle is long, and this module can be carried out algorithm work when not having line-scan digital camera.
The disclosed image processing IP core of the present invention can be applicable among the FPGA (Field-ProgrammableGate Array, field programmable gate array), and its operation result is as shown in Figure 5, and the significance signal among Fig. 5 is:
Handle in the IP kernel test process at test pattern:
Image_data (view data) is the output DATA_OUT of image measurement unit;
Image_line_valid (row useful signal) is the output LINE_VALID of image measurement unit;
In the normal use of surface quality detection system:
Image_data (view data) is the picturedeep certificate of line-scan digital camera output;
Image_line_valid (row useful signal) is the capable useful signal of line-scan digital camera output;
Pixs_sum (pixel and signal) and pixs^2_sum (pixel quadratic sum signal) are the output of image pretreatment unit pre_dsp;
As the input of image buffer storage processing unit, DSP_out_wr (data transmission enable signal) and DSP_out_RSQ (certain data characteristics value output signal) are the output signals after the image buffer storage processing unit is accomplished parallel processing together for the output of image pretreatment unit pre_dsp and DSP_wr (data processing enable signal);
In Image_index (image block call number), Image_error_data (defect image signal) and Image_block_ADDR (image block address signal) the input picture shared memory cell; Back level main equipment (like MicroBlaze or PowerPC) can and be taken data away through PLB or OPB bus convenient access.
The disclosed image processing IP core of the present invention, shown in accompanying drawing 7, for the single-frame images data, the mistiming between the view data of input and the result of output remains on the us rank at operation result among the FPGA.
The course of work of the present invention is:
1, the characteristic according to employed array camera generates the image measurement unit, so that realize the test of image processing IP core function and performance.Be without loss of generality, if the array camera that uses is industrial line-scan digital camera, can be according to its pixel and line frequency characteristic, the size of the bit wide real constant 306 of configuration image array [t, y], totalizer 305 and 309 values.
2, with reference to figure 6; The present invention confirms the image block method with FPGA resource, algorithm structure and the camera parameter comprehensively selected for use; get k behavior one two field picture of surface quality image, and a two field picture is divided into the image block of n/k piece along the direction of motion of surface quality image; Wherein, n is the pixel of line-scan digital camera.In practical application, be without loss of generality, can be set to 32 pixels by k, n is configured to 4096 pixels, and promptly continuously with the image segmentation framing, frame is divided into the image block of 128 32 * 32 pixels, and then the capacity of single image piece is 1024Byte.
3, operating personnel select the pictures different Processing Algorithm according to different demands; And image processing algorithm resolved into the submodule that is fit to the FPGA parallel processing; On System Generator platform, be implanted into image processing IP core, this image processing IP core can be gone into FPGA by directly transplanting.For the complicated module of some temporal characteristicses, can use VHDL language to write and go up and accomplish its preceding emulation and make and reach anticipated demand in Modsim, convert General Porcess Unit to through Black Box instrument.
4, call WaveScope virtual oscilloscope instrument, add the signal of being paid close attention in the hardware engineering, set the timer (time counter) in the image measurement unit; Confirmed the emulation periodicity, clicked Start Simulation and carry out emulation, with reference to Fig. 7; Verify whether each module reaches designing requirement; If discrepancy is arranged, need targetedly hardware engineering to be carried out the part adjustment, until satisfying design requirement.
5, the Compilation option in the System Generator module is arranged to Export as a pcoreto EDK; The Part option is selected the FPGA model of use; After confirming that engineering does not have design mistake; Click Generate and generate hardware identification code, the main equipment of level such as Microblaze or PowerPC call after supplying, and realize the complete of whole hardware system.
6, in hardware debug; Call ChipScope FVLA instrument; Catch the signal of being paid close attention in the hardware engineering; Post-simulation is carried out in global design to transplanting, guarantees in the patent working process that algorithm is implanted into the above-mentioned Flame Image Process hardware of this patent and realizes that the temporal characteristics and the above-mentioned course of work 4 behind the framework coincide, to reach design requirement.
7, operating personnel can be according to the changes in demand in later stage; Algorithm is optimized and adjusts; Repeat above-mentioned design cycle; Accomplish the hardware of the online detection algorithm of surface quality and realize, detect in real time that materials such as steel billet, glass, paper exist in process of production such as roll marks, be mingled with, scab, mass defects such as cut and spot.
Each embodiment adopts the mode of going forward one by one to describe in this instructions, and what each embodiment stressed all is and the difference of other embodiment that identical similar part is mutually referring to getting final product between each embodiment.For the disclosed device of embodiment, because it is corresponding with the embodiment disclosed method, so description is fairly simple, relevant part is partly explained referring to method and is got final product.
Need to prove; In this article; Term " comprises ", " comprising " or its any other variant are intended to contain comprising of nonexcludability; Thereby make to comprise that process, method, article or the equipment of a series of key elements not only comprise those key elements, but also comprise other key elements of clearly not listing, or also be included as this process, method, article or equipment intrinsic key element.Under the situation that do not having much more more restrictions, the key element that limits by statement " comprising ... ", and be not precluded within process, method, article or the equipment that comprises said key element and also have other identical element.
For the convenience of describing, be divided into various modules with function when describing above the device and describe respectively.Certainly, when embodiment of the present invention, can in same or a plurality of hardware devices, realize the function of each module.
To the above-mentioned explanation of the disclosed embodiments, make this area professional and technical personnel can realize or use the present invention.Multiple modification to these embodiment will be conspicuous concerning those skilled in the art, and defined General Principle can realize under the situation that does not break away from the spirit or scope of the present invention in other embodiments among this paper.Therefore, the present invention will can not be restricted to these embodiment shown in this paper, but will meet and principle disclosed herein and features of novelty the wideest corresponding to scope.
Claims (7)
1. the device of the online detection of a surface quality image is characterized in that, comprises image acquisition device and image processing IP core;
Said image acquisition device is used for the collection surface quality image, and said surface quality image is transferred to said image processing IP core;
Said image processing IP core is used for said surface quality image is divided into a plurality of image blocks, and parallel computation goes out the eigenwert of each said image block, and whether detect said image block according to said eigenwert qualified.
2. device according to claim 1 is characterized in that, said image processing IP core comprises image block pre-service and sheet menu unit, and the image buffer storage processing unit;
Said image block pre-service and sheet menu unit are used for the said surface quality image of pre-service, and said surface quality image are divided into a plurality of image blocks, transfer to said image buffer storage processing unit;
Whether said image buffer storage processing unit be used for the eigenwert of each said image block of parallel computation, and it is qualified to detect said image according to said eigenwert.
3. device according to claim 2 is characterized in that, said image pre-service and sheet menu unit comprise the chip selection signal generator, and the image pretreatment unit;
Said image pretreatment unit is used to realize that the pretreatment operation of said surface quality image, said pretreatment operation comprise the taking advantage of, add of view data of said surface quality image at least, and delay operation;
Said chip selection signal generator; Be used to produce chip selection signal; Said chip selection signal is for being divided into the image block of n/k piece with a two field picture along the direction of motion of said surface quality image, and said image block is transferred to the signal of said image buffer storage processing unit, wherein; N is the pixel of said image acquisition device, k behavior one two field picture of said surface quality image.
4. device according to claim 2 is characterized in that, said image buffer storage processing unit comprises image buffer storage unit and graphics processing unit;
Said image buffer storage unit is used for the said image block of buffer memory, and said image block is offered said graphics processing unit;
Said graphics processing unit; Be used to realize the eigenwert of the said image block of parallel computation; And the threshold interval of the normal surface image feature value of said eigenwert and prevision compared, if said eigenwert is not in the said threshold interval, then be judged to be the fault image block.
5. device according to claim 2 is characterized in that said image processing IP core also comprises the Image Sharing storage unit;
Said Image Sharing storage unit is used to store the image block of the eigenwert and the said fault of said image block, produces simultaneously to be used for the image block of the eigenwert of said image block and said the fault full up or half-full look-at-me to said industrial computer transmission.
6. device according to claim 5 is characterized in that, said Image Sharing storage unit comprises: eigenwert interruption generating device, eigenwert storage unit, defect image storage unit, and defect image interruption generating device;
Said eigenwert storage unit is used to store all kinds of eigenwerts that said image buffer storage processing unit calculates;
When the half of the said eigenwert of said eigenwert cell stores to himself capacity, said eigenwert interruption generating device produces the first half-full look-at-me;
When the said eigenwert of institute's eigenwert cell stores during to himself capacity whole, said eigenwert interruption generating device produces the first full up look-at-me;
Said defect image storage unit is used to store the image block of said fault;
When the data of the image block of the said fault of the said defect image cell stores half to himself capacity, said defect image interruption generating device produces the second half-full look-at-me;
When the data of the image block of the said fault of said defect image cell stores during to himself capacity whole, said defect image interruption generating device produces the second full up look-at-me.
7. according to any described device of claim 1-6, it is characterized in that said device also comprises the image measurement unit, said image measurement unit comprises image function test cell and image property test cell;
Said image function test cell, whether the eigenwert that is used to detect said image block is calculated correctly;
Said image property test cell, whether the computing velocity of eigenwert that is used to detect said image block is more than or equal to the picking rate of said surface quality image.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210227730.6A CN102759530B (en) | 2012-07-03 | 2012-07-03 | A kind of surface quality image on-line measuring device |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201210227730.6A CN102759530B (en) | 2012-07-03 | 2012-07-03 | A kind of surface quality image on-line measuring device |
Publications (2)
Publication Number | Publication Date |
---|---|
CN102759530A true CN102759530A (en) | 2012-10-31 |
CN102759530B CN102759530B (en) | 2016-05-04 |
Family
ID=47054058
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201210227730.6A Active CN102759530B (en) | 2012-07-03 | 2012-07-03 | A kind of surface quality image on-line measuring device |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102759530B (en) |
Cited By (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103344648A (en) * | 2013-07-10 | 2013-10-09 | 鞍钢股份有限公司 | Steel plate surface quality detection method and system based on GPU |
CN106770321A (en) * | 2016-12-08 | 2017-05-31 | 广东工业大学 | A kind of plastic part defect inspection method interval based on multi thresholds |
CN107402218A (en) * | 2017-09-25 | 2017-11-28 | 武汉华星光电技术有限公司 | Microdefect detection method, device and the equipment of CF substrates |
CN107767336A (en) * | 2016-08-19 | 2018-03-06 | ***通信有限公司研究院 | A kind of method and device of image procossing |
CN109632808A (en) * | 2018-12-05 | 2019-04-16 | 深圳大学 | Seamed edge defect inspection method, device, electronic equipment and storage medium |
CN109684948A (en) * | 2018-12-11 | 2019-04-26 | 歌尔股份有限公司 | The method, apparatus and system of detection object mesh exception |
CN110293566A (en) * | 2019-05-07 | 2019-10-01 | 北京鼎新时代科技有限公司 | A kind of online robot finishing platform on milling train and extended line |
CN114092748A (en) * | 2021-11-30 | 2022-02-25 | 中国科学院空天信息创新研究院 | SAR (synthetic aperture radar) unintentional interference detection method, device, equipment and medium |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1696675A (en) * | 2005-05-31 | 2005-11-16 | 哈尔滨工业大学 | Image type online flaws detection equipment in line array for surface of strip steel, and detection method |
JP2008203198A (en) * | 2007-02-22 | 2008-09-04 | Nippon Steel Corp | Surface flaw inspecting system and flaw method, and computer program |
CN101556251A (en) * | 2009-05-08 | 2009-10-14 | 西安理工大学 | CTP plate making quality testing method based on digital signal processor |
CN101988907A (en) * | 2010-09-16 | 2011-03-23 | 无锡动视科技有限公司 | Image processing system for surface detection of coiled material |
CN102253049A (en) * | 2011-06-30 | 2011-11-23 | 东北大学 | Method for accurately detecting surface quality on line in production process of band steel |
CN102509255A (en) * | 2011-10-26 | 2012-06-20 | 苏州百滨电子科技有限公司 | High-speed image acquiring and processing method and device |
-
2012
- 2012-07-03 CN CN201210227730.6A patent/CN102759530B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1696675A (en) * | 2005-05-31 | 2005-11-16 | 哈尔滨工业大学 | Image type online flaws detection equipment in line array for surface of strip steel, and detection method |
JP2008203198A (en) * | 2007-02-22 | 2008-09-04 | Nippon Steel Corp | Surface flaw inspecting system and flaw method, and computer program |
CN101556251A (en) * | 2009-05-08 | 2009-10-14 | 西安理工大学 | CTP plate making quality testing method based on digital signal processor |
CN101988907A (en) * | 2010-09-16 | 2011-03-23 | 无锡动视科技有限公司 | Image processing system for surface detection of coiled material |
CN102253049A (en) * | 2011-06-30 | 2011-11-23 | 东北大学 | Method for accurately detecting surface quality on line in production process of band steel |
CN102509255A (en) * | 2011-10-26 | 2012-06-20 | 苏州百滨电子科技有限公司 | High-speed image acquiring and processing method and device |
Non-Patent Citations (1)
Title |
---|
李长乐: ""冷轧带钢表面缺陷图像检测关键技术的研究"", 《哈尔滨工业大学工学博士学位论文》, 31 May 2011 (2011-05-31) * |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103344648A (en) * | 2013-07-10 | 2013-10-09 | 鞍钢股份有限公司 | Steel plate surface quality detection method and system based on GPU |
CN107767336A (en) * | 2016-08-19 | 2018-03-06 | ***通信有限公司研究院 | A kind of method and device of image procossing |
CN106770321A (en) * | 2016-12-08 | 2017-05-31 | 广东工业大学 | A kind of plastic part defect inspection method interval based on multi thresholds |
CN106770321B (en) * | 2016-12-08 | 2019-07-19 | 广东工业大学 | A kind of plastic part defect inspection method based on multi-threshold section |
CN107402218A (en) * | 2017-09-25 | 2017-11-28 | 武汉华星光电技术有限公司 | Microdefect detection method, device and the equipment of CF substrates |
CN109632808A (en) * | 2018-12-05 | 2019-04-16 | 深圳大学 | Seamed edge defect inspection method, device, electronic equipment and storage medium |
CN109632808B (en) * | 2018-12-05 | 2021-11-09 | 深圳大学 | Edge defect detection method and device, electronic equipment and storage medium |
CN109684948A (en) * | 2018-12-11 | 2019-04-26 | 歌尔股份有限公司 | The method, apparatus and system of detection object mesh exception |
CN109684948B (en) * | 2018-12-11 | 2021-07-27 | 歌尔股份有限公司 | Method, device and system for detecting mesh abnormality of object |
CN110293566A (en) * | 2019-05-07 | 2019-10-01 | 北京鼎新时代科技有限公司 | A kind of online robot finishing platform on milling train and extended line |
CN114092748A (en) * | 2021-11-30 | 2022-02-25 | 中国科学院空天信息创新研究院 | SAR (synthetic aperture radar) unintentional interference detection method, device, equipment and medium |
Also Published As
Publication number | Publication date |
---|---|
CN102759530B (en) | 2016-05-04 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102759530A (en) | Online detection device for surface quality image | |
CN101441271B (en) | SAR real time imaging processing device based on GPU | |
CN104198872B (en) | Online equipment for monitoring power quality and method | |
CN203133000U (en) | Image-based glass defect online detection device | |
CN103344648A (en) | Steel plate surface quality detection method and system based on GPU | |
CN104093016A (en) | Camera module smudginess detection method and system | |
CN102279386B (en) | SAR (Synthetic Aperture Radar) imaging signal processing data transposing method based on FPGA (Field Programmable Gata Array) | |
CN105208275B (en) | A kind of system for supporting to handle in real time in flow data piece | |
CN207067733U (en) | A kind of Sync image capture system based on FPGA and ARM | |
CN102509304A (en) | Intelligent optimization-based camera calibration method | |
CN114446037A (en) | Carbon emission monitoring system and method based on Internet of things technology | |
CN103823133A (en) | On-line power quality monitoring system based on compression sensing | |
CN205486305U (en) | Multinuclear heart image processor system based on compactPCI bus | |
CN101567078B (en) | Dual-bus visual processing chip architecture | |
CN103067706B (en) | Based on the IP kernel of FPGA | |
CN102446342A (en) | Reconfigurable binary arithmetical unit, reconfigurable binary image processing system and basic morphological algorithm implementation method thereof | |
CN103256975A (en) | Real-time vibration failure monitoring device of wind turbine generator system | |
CN102831572A (en) | System for collecting and processing image | |
CN102129419A (en) | Fast Fourier transform-based processor | |
CN205029751U (en) | Image data collection system based on CCD | |
CN103970904A (en) | Quick response type remote sensing big data processing system | |
CN109710562A (en) | A kind of configurable and high speed FPGA configuration circuit and implementation method based on SELECTMAP | |
CN102980527B (en) | A kind of three-dimension sensor real time high-speed data handling system based on CMOS | |
CN107767372B (en) | Chip pin online visual detection system and method for layered parallel computing | |
CN101634939B (en) | Fast addressing device and method thereof |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant |