CN204481942U - A kind of distributed transmission device for video stream signal treatment system - Google Patents

A kind of distributed transmission device for video stream signal treatment system Download PDF

Info

Publication number
CN204481942U
CN204481942U CN201520177438.7U CN201520177438U CN204481942U CN 204481942 U CN204481942 U CN 204481942U CN 201520177438 U CN201520177438 U CN 201520177438U CN 204481942 U CN204481942 U CN 204481942U
Authority
CN
China
Prior art keywords
module
video
processing unit
data
dma
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Active
Application number
CN201520177438.7U
Other languages
Chinese (zh)
Inventor
刘娟秀
张静
谢煜
倪光明
刘霖
叶玉堂
刘永
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
University of Electronic Science and Technology of China
Original Assignee
University of Electronic Science and Technology of China
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by University of Electronic Science and Technology of China filed Critical University of Electronic Science and Technology of China
Priority to CN201520177438.7U priority Critical patent/CN204481942U/en
Application granted granted Critical
Publication of CN204481942U publication Critical patent/CN204481942U/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Landscapes

  • Bus Control (AREA)

Abstract

The utility model discloses a kind of distributed transmission device for video stream signal treatment system, in video frequency signal processing embedded system, this system comprises multiple video processing unit, AXI bus, arm processor and DDR memory, distributed dma controller and corresponding DMA dispensing unit.The dma controller of each dispersion connects a video processing unit, completes its video processing unit connected to the access of DDR memory.The utility model can realize transmitting two-dimensional video data between processing unit and DDR memory, and the distributed design approach of use makes DMA control to have the advantage that speed is fast, flexibility is high, set expandability is strong, reliability is high.The utility model is applied in the processing system for video of high-speed motion cable visual defects on-line checkingi, greatly can improve the access speed of video processing unit to chip external memory DDR, thus improves the performance of whole system.

Description

A kind of distributed transmission device for video stream signal treatment system
Technical field
The utility model relates to field of data transmission, and more specifically, the utility model relates to a kind of Distributed DMA transmitting device for video stream signal treatment system.
Technical background
In process of production, because raw material are chosen and the many factors such as processing technology, inevitably can there is the defects such as pore, cut, pit in wire surface to wire rod, these defects can affect the performance of final products.Therefore need to detect its visual defects in the process of wire production.Mechanical vision inspection technology is the image utilizing video camera to obtain detected object, is realized a kind of new and high technology detected, advantages such as having noncontact, speed is fast, precision is high, be intelligent by image processing techniques.Machine vision technique is incorporated in the detection of wire rod visual defects, replaces traditional manual detection, overcome manual detection poor stability, slow-footed shortcoming.
Video algorithm processing section is the core of machine vision.Because cable has very high movement velocity, and defects detection has very high required precision, thus video data is caused to have the feature of high-resolution and high frame per second, in order to ensure the real-time of process, needs on FPGA, use hardware logic resource and matching embedded type arm processor completes video processing function jointly, wherein use the hardware logic resource of FPGA to complete some hardware video processing units, complete the single step in video processing procedure.
In video algorithm treatment system, comprise arm processor, each hardware video processing unit, jumbo external memory storage and corresponding interface and control module.In the process of processing system for video work, arm processor and each hardware video processing unit need frequently, externally write and read data in memory DDR in large quantities.Thus, the transmission rate of video data between processing unit and DDR becomes the key factor affecting whole processing system for video.
DMA (Direct Memory Access direct memory access) is a kind of data transfer mode of high speed, and the intervention without the need to CPU can realize direct read/write data between peripheral hardware and memory.The utility model discloses a kind of Distributed DMA transmission method for video stream signal treatment system and device, for the video algorithm treatment system that cable visual defects detects, use the transfer of data between dma mode controlled processing unit and DDR, greatly can improve the access speed of video processing unit to chip external memory DDR, thus improve the performance of whole system.
1, general dma controller, the one-dimensional data of most transmission, and video data is the view data of two dimension, and there is the proprietary marking signal of video data, the marking signal that traditional DMA control method cannot utilize video data special, directly cannot realize the transmission of two-dimensional video signal.
2, centralized DMA control mode, uses a multichannel dma controller, is all connected to by all peripheral hardwares on this dma controller, unified data carrying work by this dma controller.When the method is applied in the FPGA video signal processing field of the utility model use, there is following shortcoming:
(1) system flexibility is poor, and the mode of different peripheral hardware transmission data also exists certain difference, needs the pattern adopting different transmission, uses centralized DMA control mode, be difficult to meet this requirement.
(2) set expandability is poor, in time having new peripheral hardware to need to add system to, uses centralized DMA control method, needs to remodify dma controller, increase the designs such as corresponding passage and arbitration, very inconvenient in actual applications.
(3) reliability is low, and when transmission fault appears in certain peripheral hardware, concentrated dma controller can be affected, and may cause normally working, thus causes other peripheral hardwares also cannot by DMA normal transmission data, and whole transmission system will be collapsed.
Summary of the invention
Technical problem to be solved in the utility model is a kind of distributed transmission device for video stream signal treatment system of weak point design for background technology, processed step by step needing video stream signal to be processed by hardware, each treatment step adopts a dma controller to access this step data separately, thus reaches in data access process the object that reliability is high, speed is fast, flexibility is strong, set expandability is strong.
A kind of distributed transmission device for video stream signal treatment system of the utility model, this device carries out distributed treatment to video stream signal, each treatment step passes through hardware implementing, often walk processing hardware and all can control by DMA dispensing unit the read-write that a dma controller is responsible for this step processing hardware desired data, thus realize goal of the invention.Thus a kind of distributed transmission device for video stream signal treatment system of the utility model, this device comprises: multiple signal processing module, arm processor module, memory module and connect the transfer bus of each module, and each signal processing module is responsible for process signal being carried out to a step; Wherein signal processing module comprises: video processing unit, dma controller, DMA dispensing unit, DMA dispensing unit judges the reading and writing data demand of video processing unit, controls dma controller by video processing unit desired data direct read/write from memory module;
Described dma controller comprises: receiver module, write access cache module, sending module, read channel cache module, Read-write Catrol module, control command receiver module, control module; Wherein control command receiver module receives the order of DMA dispensing unit, is then transferred to control module, and control module carries out work according to order control receiver module, sending module, Read-write Catrol module; The data of write memory module are needed, then stored in write access cache module, then by read-write controller by writing data into memory module by receiver module receiver, video processing unit; By Read-write Catrol module sense data from memory module, the data write read channel cache module then will read, then send data to video processing unit by sending module.
In wherein said signal processing module, video processing unit is connected with 1 dma controller, and its reading and writing data are completed by a dma controller.
In wherein said signal processing module, video processing unit is connected with 2 dma controllers, and its read data is completed by a dma controller, writes data and is completed by another dma controller; Described dma controller is connected with 2 video processing units, controls the read data of 1 data cell respectively, another data cell write data.
Wherein said transfer bus adopts AXL4 bus, and described video processing unit, interface between DMA dispensing unit and dma controller all adopt AXI-stream agreement.
A kind of distributed transmission device for video stream signal treatment system of the utility model, the DMA of the whole device of this utility model controls to be distributed to each video processing unit, each video processing unit connects a DMA, and the DMA of each dispersion has been responsible for its video processing unit connected to the access of memory; This utility model tool has the following advantages:
1. distributed system has speed faster, and the DMA concurrent working of each distribution, can have speed faster than single centralized DMA;
2. the DMA of each distribution according to the processing demands of different video processing units, can be configured to different working methods, has very strong flexibility;
3. in time having new video processing unit to need to add native system to, use Distributed DMA data transmission system, the dma controller that can increase a distribution is connected to newly-increased video processing unit, the expansion of completion system very easily;
4., when dma controller or certain video processing unit appearance exception of certain dispersion, the normal work that the DMA that can not affect other controls, thus system has very high reliability.
Accompanying drawing explanation
Fig. 1 is the schematic diagram of a kind of distributed transmission device for video stream signal treatment system of the utility model.
Specific embodiment
1, for a Distributed DMA transmitting device for video stream signal treatment system, be applied in video frequency signal processing embedded system.
2, described video frequency signal processing embedded system, as shown in Figure 1, comprises video processing unit (video processing unit 1, video processing unit 2 ..., video processing unit 2N), DMA dispensing unit (DMA dispensing unit 1, DMA dispensing unit 2 ... DMA dispensing unit N), dma controller (dma controller 1, dma controller 2, dma controller N), AXI bus, arm processor and DDR memory.
3, described video processing unit (video processing unit 1, video processing unit 2, video processing unit 2N), refer to some video signal processing module, each unit completes certain particular step in whole video processnig algorithms, and described processing unit, by the data in distributed dma controller read-write DDR memory described in the utility model, requires that the connecting interface of video processing unit and dma controller is AXI-stream agreement.
4, described DMA dispensing unit (DMA dispensing unit 1, DMA dispensing unit 2, DMA dispensing unit N), refer to the module of configuration DMA working method, there is AXI-lite interface, for AXI-lite main frame, the dma controller of each dispersion uses a DMA dispensing unit to configure its mode of operation, the information of DMA dispensing unit configuration comprises that startup is ceased and desisted order, picture altitude information, picture traverse information and initial address message (IAM), these signals have independently two groups, respectively corresponding write access and read channel.
5, described AXI bus is the bus that system uses, and connects modules.
6, described arm processor is the embedded-type ARM stone that FPGA is embedded, the overall control of completion system and the work of partial video Processing Algorithm.
7, described DDR memory, be the outer Large Copacity DDR memory of FPGA sheet, have the advantage that capacity is large, read or write speed is fast, function is that buffered video data is for algorithm unit process.
8, the described Distributed DMA transmission method for video stream signal treatment system and device, described distributivity show that the realization of dma controller is separated into dma controller 1, dma controller 2 ..., dma controller N.The dma controller of each dispersion connects corresponding video processing unit, completes its video processing unit connected to the access of DDR memory.
9, described distribution dma controller (i.e. dma controller 1, dma controller 2 ..., dma controller N), comprise independently passage write access and read channel, and corresponding control section.FPGA realizes, completes its video processing unit connected to the access of DDR memory.The independence of described write access and read channel, refers to that write access and read channel work alone, does not affect each other.
The dma controller of described distribution and the connection of video processing unit have very large flexibility, can configure flexibly according to actual treatment demand.The dma controller of each distribution comprises a write access and a read channel, and these two passages can connect same video processing unit, also can connect different video processing units respectively.As shown in Figure 1 for connecting different video processing units, other connections in like manner.
10, described write access, comprises FIFO cache module on AXI-stream receiver module and write access sheet.Function is the video stream data receiving AXI-stream agreement, effective video data is extracted, is cached to FIFO cache module on write access sheet, for other modules.
11, described AXI-stream receiver module, function is the video stream signal receiving AXI-stream agreement, extracts effective video data according to corresponding marking signal, and to be stored on write access sheet in FIFO cache module.Described marking signal comprises two parts: the video format order that (1) described control section sends, and video format comprises length and the width information of a two field picture; (2) comprise frame start signal and row end signal in AXI-stream video flowing, frame start signal refers to the commencing signal of a frame video image, and row end signal refers to the marking signal that a line view data terminates.
12, FIFO cache module on described write access sheet, uses storage resources on FPGA sheet to realize, completes the buffer memory of video data in write access.Use this module can isolate video processing unit and bus transfer, data in advance after video processing unit process is written in this module, after write data volume reaches some, described Read-write Catrol module adopts burst mode the data in this module to be written in DDR memory by AXI bus.
13, described read channel, comprises FIFO cache module and AXI-stream sending module on read channel sheet.Function is the video data that buffer memory reads from DDR memory, when video processing unit needs deal with data, the data of buffer memory is sent to video processing unit according to AXI-stream agreement.
14, FIFO cache module on described read channel sheet, uses storage resources on FPGA sheet to realize, completes the buffer memory of video data in read channel.Use this module can isolate bus transfer and video processing unit, the video data of assigned address in DDR memory is cached to this module in advance, when video processing unit needs processing video data, the video data in this module reads out and sends to video processing unit by described AXI-stream sending module.
15, described AXI-stream sending module, function is reading video data in FIFO cache module from described read channel sheet, according to the signal that described control section sends, add marking signal to data, send vision signal to video processing unit with AXI-stream agreement.The signal that described control section sends is video format order, and video format comprises length and the width information of a two field picture.Described marking signal comprises frame start signal and row end signal, and frame start signal refers to the commencing signal of a frame video image, and row end signal refers to the marking signal that a line view data terminates.
16, described control section, comprises control command receiver module, control module and Read-write Catrol module.Function is receive the control information that DMA control unit sends, and according to this control information, the mode of operation of configurable write passage and read channel, and controls the mutual of described write access and read channel and AXI bus.
17, described control command receiver module, has AXI-lite bus interface, for AXI-lite is from machine, receives the control information that DMA dispensing unit sends, and this control information is written in corresponding register.Described control information comprises that startup is ceased and desisted order, picture altitude information, picture traverse information and initial address message (IAM), and these signals have independently two groups, and write access and read channel use one group respectively.Described two groups of startups are ceased and desisted order, and control startup and the stopping of write access and read channel respectively.Described two groups of picture altitude information and picture traverse information, be respectively the height of write access and read channel transmission of video, width information.Described two groups of initial address message (IAM)s, are respectively the initial address of a frame video image in DDR memory of write access and read channel.
18, described control module, function is the control information received according to described control command receiver module, controls the work of described Read-write Catrol module, AXI-stream receiver module and AXI-stream sending module.Above-mentioned control information comprises that startup is ceased and desisted order, picture altitude information, picture traverse information and initial address message (IAM).
19, described Read-write Catrol module, function is the control information receiving the transmission of described control module, calculate corresponding go out corresponding address, data in FIFO cache module on write access sheet are written in DDR memory assigned address by AXI bus, the video data in assigned address in DDR memory are write on read channel sheet in FIFO cache module.Picture altitude information, picture traverse information and initial address message (IAM) that the calculating of address above mentioned uses control module to send.This module is Time Slice Circular Scheduling algorithm to the scheduling strategy of read/write channel.

Claims (4)

1. the distributed transmission device for video stream signal treatment system, this device comprises: multiple signal processing module, arm processor module, memory module and connect the transfer bus of each module, and each signal processing module is responsible for process signal being carried out to a step; Wherein signal processing module comprises: video processing unit, dma controller, DMA dispensing unit, DMA dispensing unit judges the reading and writing data demand of video processing unit, controls dma controller by video processing unit desired data direct read/write from memory module;
Described dma controller comprises: receiver module, write access cache module, sending module, read channel cache module, Read-write Catrol module, control command receiver module, control module; Wherein control command receiver module receives the order of DMA dispensing unit, is then transferred to control module, and control module carries out work according to order control receiver module, sending module, Read-write Catrol module; The data of write memory module are needed, then stored in write access cache module, then by read-write controller by writing data into memory module by receiver module receiver, video processing unit; By Read-write Catrol module sense data from memory module, the data write read channel cache module then will read, then send data to video processing unit by sending module.
2. a kind of distributed transmission device for video stream signal treatment system as claimed in claim 1, it is characterized in that in described signal processing module, video processing unit is connected with 1 dma controller, its reading and writing data are completed by a dma controller.
3. a kind of distributed transmission device for video stream signal treatment system as claimed in claim 1, it is characterized in that in described signal processing module, video processing unit is connected with 2 dma controllers, its read data is completed by a dma controller, writes data and is completed by another dma controller; Described dma controller is connected with 2 video processing units, controls the read data of 1 data cell respectively, another data cell write data.
4. a kind of distributed transmission device for video stream signal treatment system as claimed in claim 1, it is characterized in that described transfer bus adopts AXL4 bus, described video processing unit, interface between DMA dispensing unit and dma controller all adopt AXI-stream agreement.
CN201520177438.7U 2015-03-27 2015-03-27 A kind of distributed transmission device for video stream signal treatment system Active CN204481942U (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201520177438.7U CN204481942U (en) 2015-03-27 2015-03-27 A kind of distributed transmission device for video stream signal treatment system

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201520177438.7U CN204481942U (en) 2015-03-27 2015-03-27 A kind of distributed transmission device for video stream signal treatment system

Publications (1)

Publication Number Publication Date
CN204481942U true CN204481942U (en) 2015-07-15

Family

ID=53637711

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201520177438.7U Active CN204481942U (en) 2015-03-27 2015-03-27 A kind of distributed transmission device for video stream signal treatment system

Country Status (1)

Country Link
CN (1) CN204481942U (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104717433A (en) * 2015-03-27 2015-06-17 电子科技大学 Distributed transmission device for video stream signal processing system
WO2020047843A1 (en) * 2018-09-07 2020-03-12 深圳鲲云信息科技有限公司 Method and system for transmitting video image data, storage medium, and program product

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104717433A (en) * 2015-03-27 2015-06-17 电子科技大学 Distributed transmission device for video stream signal processing system
WO2020047843A1 (en) * 2018-09-07 2020-03-12 深圳鲲云信息科技有限公司 Method and system for transmitting video image data, storage medium, and program product

Similar Documents

Publication Publication Date Title
CN208971624U (en) Automotive camera system
CN108712626B (en) Multichannel image acquisition system
US7895375B2 (en) Data transfer apparatus and data transfer method
CN102495920B (en) Integrated logic analysis module based on PCIe (peripheral component interconnection express) for FPGA (field programmable gate array)
CN108092753B (en) Hot standby redundancy system adopting hardware memory moving synchronization
CN105573951B (en) A kind of ahb bus interface system for data stream transmitting
CN104851069A (en) Cable apparent image defect detection device
CN1952979A (en) Multiple graphics processor systems and methods
CN102202171A (en) Embedded high-speed multi-channel image acquisition and storage system
CN105549916A (en) Peripheral component interconnect express (PCIe) solid state hard disk controller, PCIe based storage system and data read and write method thereof
CN204481942U (en) A kind of distributed transmission device for video stream signal treatment system
CN103593315A (en) Direct multi-hard-disk high-speed parallel reading and writing method based on FPGA
CN104811643A (en) Image data high speed storage system based on SD card array
CN106155971A (en) The referee method of I2C bus and device
CN104133802A (en) Data exchange method between Vxworks embedded operating system and motion control card of double-workpiece-platform scanning mask aligner
CN111506249A (en) Data interaction system and method based on ZYNQ platform
CN103077139B (en) Use integrated circuit and the control method thereof of internal integrate circuit bus
CN116225990A (en) Multichannel DDR read-write arbitration device based on FPGA
CN107127811A (en) Flexible material cutting robot intelligent digital controller and implementation method
CN104717433A (en) Distributed transmission device for video stream signal processing system
CN106776403A (en) A kind of high-speed high capacity storage system and its implementation based on FPGA
DE102010008907A1 (en) USB host controller and control method for USB host controller
CN103309828B (en) A kind of SD card is from controller and control method
JP2011065587A (en) Processing system and testing device
CN105744120A (en) High speed data collection card and data collection method

Legal Events

Date Code Title Description
C14 Grant of patent or utility model
GR01 Patent grant