CQ Publishing FPGA magazine No. 15

Image processing is a typical field where FPGAs are often used. In order to realize higher-performance and higher-quality products, image recognition functions are required as more intelligent functions, not just image processing. In the field of automobiles, automatic braking that has already been put into practical use and is widely used, and those equipped with conditional but advanced automatic driving functions have appeared. In these automatic driving assistance systems ADAS (Advanced Driver Assistance Systems), it is necessary to recognize white lines on the side of the road from images and identify pedestrians. FPGA is the royal road to perform these image recognition processes at high speed. In this special issue, we will build a real-time image processing system that detects a person from an input image using the results of machine learning of a person image on a PC. In addition, we will apply the difference detection technology of video frames to manufacture a sensor camera combined with infrared detection and a lane departure warning device that detects the center line of the road and issues a warning when it goes out of the lane. Table of Contents Special Feature: Agriculture in Cars !? Challenges in the AI Era! Real-Time Video Recognition Prologue Various Image Recognition Technologies Used in Actual Products Chapter 1 Implementation of Real-Time Image Processing System Chapter 2 Start Recording by Infrared Detection + Motion Detection Manufacture of sensors and cameras Chapter 3 Making lane departure warning devices with FPGA + CMOS camera modules Appendix HDMI images and camera images! Anything image processing board CQBB-IMG9 Chapter 4 Calculate the total amount from input images! MATLAB / FPGA Prototyping with Simulink High-level synthesis tool Vivado HLS Special article Chapter 1 Let's develop FPGA in C language using high-level synthesis tool Vivado HLS! Chapter 2 AXI4-How to make a multiplying IP core that can be connected to the Lite bus Chapter 3 AXI4 -Lite version Multiplying IP core is controlled from ARM processor Latest technology explanation Porting of BSP necessary for developing FPGA with Altera SDK for OpenCL Implementation of handwritten number recognition by multi-layer neural network (Introduction) FPGA implementation from MIPS head family Introducing the latest device "MIPSfpga" for the core "MIPSfpga" Overview of programmable analog IC analogram