Cartesian Genetic Programming for Image Processing

Harding, Simon; Leitner, Jürgen; Schmidhuber, Jürgen

doi:10.1007/978-1-4614-6846-2_3

Cited by 51 publications

(29 citation statements)

References 21 publications

(20 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Sample supervised segment generation can be seen as an explicit example of more general approaches found at the intersection of evolutionary computation and image analysis/computer vision [19]. A distinction can be made [14] based on the granularity of the search process-whether the search method is used to construct a segmentation algorithm/image processing method [20][21][22][23][24], common with cellular automata, mathematical morphology and genetic programming approaches, or either for tuning the free parameters of an algorithm [14,[25][26][27][28].…”

Section: Sample Supervised Segment Generationmentioning

confidence: 99%

Data Transformation Functions for Expanded Search Spaces in Geographic Sample Supervised Segment Generation

Fourie

Schöepfer

2014

Remote Sensing

View full text Add to dashboard Cite

Abstract:Sample supervised image analysis, in particular sample supervised segment generation, shows promise as a methodological avenue applicable within Geographic Object-Based Image Analysis (GEOBIA). Segmentation is acknowledged as a constituent component within typically expansive image analysis processes. A general extension to the basic formulation of an empirical discrepancy measure directed segmentation algorithm parameter tuning approach is proposed. An expanded search landscape is defined, consisting not only of the segmentation algorithm parameters, but also of low-level, parameterized image processing functions. Such higher dimensional search landscapes potentially allow for achieving better segmentation accuracies. The proposed method is tested with a range of low-level image transformation functions and two segmentation algorithms. The general effectiveness of such an approach is demonstrated compared to a variant only optimising segmentation algorithm parameters. Further, it is shown that the resultant search landscapes obtained from combining mid-and low-level image processing parameter domains, in our problem contexts, are sufficiently complex to warrant the use of population based stochastic search methods. Interdependencies of these two parameter domains are also demonstrated, necessitating simultaneous optimization.

show abstract

Section: Sample Supervised Segment Generationmentioning

confidence: 99%

Data Transformation Functions for Expanded Search Spaces in Geographic Sample Supervised Segment Generation

Fourie

Schöepfer

2014

Remote Sensing

View full text Add to dashboard Cite

show abstract

“…This requires consideration of the elementary building blocks of ADTs, viz. products and coproducts 6 . The conversion of an ADT to and from this representation is described extensively by Hinze [24] and is beyond the scope of this article, but fortunately the Scala library Shapeless [25,26] provides complete support for this and a variety of other polytypic methods (e.g.…”

Section: Product and Coproduct Typesmentioning

confidence: 99%

“…Technically, this is often achieved by mapping the host language API of interest to individual functions in the GP instruction set. For instance, in order to manipulate programs which use the API of a computer vision library (e.g., OpenCV [3], as in the GP/GI work of [4,5,6]), one could provide adaptor code for each API call of interest. This task is nowadays greatly facilitated by availability of a rich choice of domain-agnostic software packages (ECJ [2], EpochX [7] and DEAP [8] to name a few), which offer a extensive support for the representations and operators of GP.…”

Section: Introductionmentioning

confidence: 99%

Polytypic Genetic Programming

Swan

Krawiec

Ghani

2017

Applications of Evolutionary Computation

View full text Add to dashboard Cite

Abstract. Program synthesis via heuristic search often requires a great deal of 'boilerplate' code to adapt program APIs to the search mechanism. In addition, the majority of existing approaches are not type-safe: i.e. they can fail at runtime because the search mechanisms lack the strict type information often available to the compiler. In this article, we describe Polytope, a Scala framework that uses polytypic programming, a relatively recent advance in program abstraction. Polytope requires a minimum of boilerplate code and supports a form of strongtyping in which type rules are automatically enforced by the compiler, even for search operations such as mutation which are applied at runtime. By operating directly on language-native expressions, it provides an embeddable optimization procedure for existing code. We give a tutorial example of the specific polytypic approach we adopt and compare both runtime efficiency and required lines of code against the well-known EpochX GP framework, showing comparable performance in the former and the complete elimination of boilerplate for the latter.

show abstract

“…A comprehensive review of CGP can be found in [20]. In this work, we use a version called 'CGP for Image Processing' (CGP-IP) [21]. Previously CGP-IP has been applied to problems such as medical imaging, robot vision, terrain classification [22] and image filters.…”

Section: Learning Object Identification Using Cartesian Genetic Pmentioning

confidence: 99%

“…In this work, the function set comprises a large subset of the OpenCV image processing library. Over 60 functions are available, and a complete list can be found in [21]. This method enables inserting of domain knowledge into evolved programs, and also improves evolvability as operations that are useful can be used directly instead of re-evolving the same functionality.…”

Section: Learning Object Identification Using Cartesian Genetic Pmentioning

confidence: 99%

Autonomous learning of robust visual object detection and identification on a humanoid

Leitner

Chandrashekhariah

Harding

et al. 2012

2012 IEEE International Conference on Development and Learning and Epigenetic Robotics (ICDL)

Self Cite

View full text Add to dashboard Cite

Abstract-In this work we introduce a technique for a humanoid robot to autonomously learn the representations of objects within its visual environment. Our approach involves an attention mechanism in association with feature based segmentation that explores the environment and provides object samples for training. These samples are learned for further object identification using Cartesian Genetic Programming (CGP). The learned identification is able to provide robust and fast segmentation of the objects, without using features. We showcase our system and its performance on the iCub humanoid robot.

show abstract

Cartesian Genetic Programming for Image Processing

Cited by 51 publications

References 21 publications

Data Transformation Functions for Expanded Search Spaces in Geographic Sample Supervised Segment Generation

Data Transformation Functions for Expanded Search Spaces in Geographic Sample Supervised Segment Generation

Polytypic Genetic Programming

Autonomous learning of robust visual object detection and identification on a humanoid

Contact Info

Product

Resources

About