Object detection in multi-modal images using genetic programming

Bhanu, Bir; Lin, Yingqiang

doi:10.1016/j.asoc.2004.01.004

Cited by 46 publications

(30 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Suppose G and G′ are foregrounds in the ground truth image and the resultant image of the composite operator respectively. Let n(X) denote the number of pixels within region X, then Fitness = n (G∩G′) / n (G  G′) [2,4].…”

Section: Fitness Valuementioning

confidence: 99%

“…c) Mutation: mutation is introduced to randomly change the structure of some individuals. We are using tree type of mutation [2,3]:…”

Section: E Operations Of Genetic Programmingmentioning

confidence: 99%

“…The search is done by performing selection, crossover and mutation operations in GP [2,3]. a) Selection: genetic operators in GP are applied to individuals that are probabilistically selected based on fitness.…”

Section: E Operations Of Genetic Programmingmentioning

confidence: 99%

“…Object recognition algorithms rely on matching or learning algorithms using appearance-based or feature-based techniques. We are trying to achieve good recognition results using the feature-based technique [2,6]. The quality of object recognition is heavily dependent on the effectiveness of features.…”

Section: Introductionmentioning

confidence: 99%

“…Also, the inherent parallelism of GP and the speed of computers allows a much larger portion of the search space to be explored. We have used a simple (steadystate) [2] genetic programming algorithm and primitive features to detect and to recognize objects in SAR images. A program system is developed based on this GP algorithm by means of which two experiments were done: trying to recognize five types of objects, trying to recognize seven types of objects.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Recognition of Objects by Using Genetic Programming

Safaryan¹,

Sarukhanyan²

2013

IJACSA

View full text Add to dashboard Cite

Abstract-This document is devoted to the task of object detection and recognition in digital images by using genetic programming. The goal was to improve and simplify existing approaches. The detection and recognition are achieved by means of extracting the features. A genetic program is used to extract and classify features of objects. Simple features and primitive operators are processed in genetic programming operations. We are trying to detect and to recognize objects in SAR images. Due to the new approach described in this article, five and seven types of objects were recognized with good recognition results. Keywords-Terminals;Fitness; Selection; Crossover; Mutation; Ground Truth I. INTRODUCTION Object recognition is still a challenge for computer vision systems in general. The main purpose of object recognition is to identify the kinds of the objects in an image [1]. Object recognition algorithms rely on matching or learning algorithms using appearance-based or feature-based techniques. We are trying to achieve good recognition results using the feature-based technique [2,6]. The quality of object recognition is heavily dependent on the effectiveness of features. The features used to represent an object are the key to the object detection and recognition. It is difficult to extract good features from real images due to various factors, including noise. There are many features that can be extracted. It is very difficult to find appropriate features and to synthesize composite features. Synthesizing effective new features from primitive features is equivalent to finding good points in the feature combination space where each point represents a combination of primitive features. The feature combination space and feature subset space are huge and complicated and it is very difficult to find good points in such vast spaces unless one has an efficient search algorithm [5,7]. Genetic programming (GP) is used as search algorithm. GP may try many unconventional combinations and in some cases these unconventional combinations yield exceptionally good recognition performance. Also, the inherent parallelism of GP and the speed of computers allows a much larger portion of the search space to be explored. We have used a simple (steadystate) [2] genetic programming algorithm and primitive features to detect and to recognize objects in SAR images. A program system is developed based on this GP algorithm by means of which two experiments were done: trying to recognize five types of objects, trying to recognize seven types of objects.

show abstract

Section: Fitness Valuementioning

confidence: 99%

“…c) Mutation: mutation is introduced to randomly change the structure of some individuals. We are using tree type of mutation [2,3]:…”

Section: E Operations Of Genetic Programmingmentioning

confidence: 99%

Section: E Operations Of Genetic Programmingmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%