A Bottom-up Approach for Visual Object Recognition on FPGA based Embedded Multiprocessor Architecture

Hanen Chenini

doi:10.14569/ijacsa.2017.080557

Abstract

This paper presents an object recognition approach of outdoor autonomous systems identifying the nature of the interested object when observing an image. Therefore, seeking for effective and robust recognition method, the proposed approach is performed using a novel saliency based feature detector/descriptor which is combined with an object classifier to identify the nature of objects in an indoor or an outdoor environment. As known, bottom-up visual attention computational models need a considerable computational power and communication cost. A major challenge in this work is to deal with such image processing applications managing a large amount of the information processing and to work within real-time requirements by improving the processing speed. Based on interesting approach designing specific architectures for parallelism, this paper presents a solution for rapid prototyping of saliency-based object recognition applications. In order to meet computation and communication requirement, the developed pipelined architectures are composed of identical processing modules which can work concurrently with distributed memories and compute in parallel several sequential tasks with a high computational cost. We present hardware implementations with performance results on an Xilinx System-on-Programmable Chip (SoPC) target. The experimental results including execution times and application speedups as well as requirements in terms of computing resources show that the proposed homogeneous network of processors is efficient for embedding the proposed image processing application.

Highlights

Object recognition in autonomous systems is an important task in building a system that can sense, identify the nature of objects around it and afterward react according to this information
Attentional models of the input image are compared with the trained attentional model and the the current feature model giving maximum correspondence is considered the best match of the target object
We perform experiments demonstrating the properties of our object recognition approach and second we provide experimental results of the pipelined architecture implemented on a Virtex6-LX760T Field Programmable Gate Arrays (FPGA) and compare its performance with two existing HMAX accelerators tailored to saliency based object recognition algorithm

Summary

Introduction

Object recognition in autonomous systems (robots, vehicles, UAVs, etc.) is an important task in building a system that can sense, identify the nature of objects around it and afterward react according to this information (exploring unknown environments, obstacle avoiding, computing flight paths, etc.). In order to apply saliency for object recognition, we need to obtain the saliency maps for three distinct features (color, intensity, and orientation). As a result, this method yields an output map containing only the regions that constitute the most salient regions. To guide the attention to look for reference objects, each saliency model is classified as container or as non-container of each reference model by computing a dissimilarity score between each extracted model (current object model) and each target model (reference model) via a matching process. We can eliminate the salient regions that don’t contain the target objects, and the result can be used in segmenting the whole color image. The image processing applications based saliency computations are naturally www.ijacsa.thesai.org

Objectives

Results

Conclusion