A robot vision system module for analysis and design (1)

First, an overview

Vision technology is in recent decades the development of an emerging technology.

Machine vision can replace the Visual engaged in inspection of mankind, target tracking, robot guidance, and other aspects of work, especially those requiring repeat, rapid accurate from the image gets information. Although the current hardware and software technology, and machine vision functions are still at the primary level, but its potential value attracted world attention to developed countries like the United States, Japan, Germany, France, have put a lot of manpower and material resources to conduct research, in recent years has been in some aspects of machine vision obtained breakthrough progress, machine vision in vehicle safety technology, automation technology, and other applications are increasingly showing their key value. This article according to the latest CMOS image capture chip design a common vision system module, after compiling different image processing, pattern recognition algorithm is this module can be applied to the robot soccer, unmanned vehicles, and so on various occasions.

2. design principles

System block diagram shown in Figure 1.

The system consists of five main chip: image acquisition chip OV7620, high-speed microprocessor SH4, large-scale FPGA programmable arrays, and serial port control IC MAX232.

FPGA internal programming establish two dual-port RAM, produce the image sensor the desired frequency, the line field synchronization signals, as well as the control of dual-port RAM storage time series. SH4 responsible for configuration via I2C OV7620, read dual-port RAM image data to be processed, and information through a serial port image upload or control stepper motors and other equipment.

3. image acquisition module

System module for CMOS image sensor OV7620 as core, also including a condenser head and other auxiliary components such as the crystal oscillator, resistance 27MHZ capacitors, etc.