Video Enhancement for ADAS Systems based on FPGA and CNN Platform
نویسندگان
چکیده
In road-transportation terminology, ADAS (Advanced Driver Assistance System) is a combination of monitoring-, alarm-, and control systems for increasing safety level in vehicles. Most of ADAS systems are infrastructureindependent and autonomous. This means these systems must work without any external awareness systems on the roads during the drive. Nevertheless, there are some systems which can cooperate with other vehicles (i.e. CACC Cooperative Adaptive Cruise Control) or awareness systems on the road. Camera is the main part of monitoring and observation unit in ADAS systems. In some cases, fusion of camera data (video frames) and LIDAR (Light Detection And Ranging) leads to more safety and confidence in data. Most of ADAS sub systems such as Adaptive Cruise Control (ACC) , Lane Departure Warning (LDW) , Automatic Parking, Collision Warning Avoidance (CWA) systems and Lane Change Assistance are based on video information. Enhancing video frames is possible by common image processing techniques. These classical algorithms with the sequential processing architecture unit take lot of resources and energy. For producers of ADAS, the main challenges for enhancing image quality are speed and manufacturing cost. It is possible to overcome these problems by developing a proper framework and architecture. In this research, we propose a new method of CNN design based on ESL (Electronic System Design). To implement this model we used fixed-point arithmetic technique and discrete model for CNN.
منابع مشابه
Design Enhancement of Combinational Neural Networks Using Hdl Based Fpga Framework for Pattern Recognition
The fast emerging highly-integrated multimedia devices require complex video/image processing tasks leading to a very challenging design process; as it demands more efficient and high processing systems. Neural networks are used in many of these imaging applications to represent the complex input-output relationships. Software implementation of these networks attain accuracy with tradeoffs betw...
متن کاملFPGA-based of Thermogram Enhancement Algorithm for Non-destructive Thermal Characterization
متن کامل
Embedded system for contrast enhancement in low-vision
This paper presents a real-time contrast enhancement system, implemented in FPGA and adapted to display the processed images on a Head Mounted Display (HMD). A novel visual processing scheme is proposed which combines a version of the algorithm known as Contrast Limited Adaptive Histogram Equalization (CLAHE) with a spatial filtering based on a bio-inspired retina model. The system is designed ...
متن کاملPlatform-Based Design Approach for Embedded Vision Applications
Implementation of an efficient, complex real-time image and video processing system has many different contrary design needs. These systems require managing and processing a large amount of data in a timely manner. The robust implementation requires design reuse and flexibility, which is very crucial for an ever-increasing application need. A combination of hardware and software which form a pl...
متن کاملUsing Reconfigurable Supercomputers and C-to-Hardware Synthesis for CNN Emulation
The complexity of hardware design methodologies represents a significant difficulty for non hardware focused scientists working on CNN-based applications. An emerging generation of Electronic System Level (ESL) design tools is been developed, which allow software-hardware codesign and partitioning of complex algorithms from High Level Language (HLL) descriptions. These tools, together with High...
متن کامل