The Logic Interface Design Of Smart Camera And The Implementation Of ROI

With the continuous advancement and development of science and technology,modern industry is entering the era of intelligence,and the use of smart cameras in industrial production is becoming more and more extensive,such as bar code recognition,character recognition,high-precision dimensional measurement of industrial parts,etc.Related technologies research of smart camera includes smart camera startup mode,the image sensor and related device drivers,on-chip image algorithms acceleration.Smart cameras can perform various functions such as image acquisition,processing,and communication,and have high requirements for software and hardware.Around the intelligent camera logic interface design and ROI function implementation,it choose to complete the design development on the ZYNQ7000 platform of ARM+FPGA architecture.This article mainly studies the following aspects:(1)Design logic interface,based on industrial image sensors Python300 and 2000,complete the raw image output from the sensor to differential single-ended,high-speed serial-to-parallel conversion,synchronous channel decoding,multi-channel integration and timing synchronization to make the original image Into video format data;(2)The data flow in the image acquisition path is studied.The dependence of ROI on image sensor,logic interface and subsequent modules of logic interface is analyzed.The internal control mode of ARM module on each IP module of FPGA is described,and ROI parameters are pointed out.How to adapt the working mode of each module in the image acquisition path to the new ROI when changing;(3)Based on Python2000,the strip-shaped ROI function is implemented,that is,the ROI size and position are specified,which can make the acquisition path transmit the image data of the set area,and compare and analyze the relationship between the frame rate and the original frame rate under different ROI sizes.Experiments show that the logic interface module can effectively solve the difference between the image sensor Python2000 and Python300 raw data,serial-to-parallel conversion,remapping,multi-channel integration and timing synchronization,and the image acquisition path can output a clear image with stable frame rate;ROI function The acquisition path can only transmit the image data of the set area,and the frame rate is improved.Research has practical value and can be adapted to many occasions of image acquisition and processing,and has certain significance in academic research and engineering applications.