Evolutionary Computer Vision (ECV) is at the intersection of two major research fields of artificial intelligence: 1) computer vision (CV) and 2) evolutionary computation (EC). This special issue brings an overview of state-of-the-art contributions to the latest research and development in the discipline. CV includes methods for acquiring, processing, analyzing, and understanding images. The aim is to design computational models of human and animal perception. ECV is an interdisciplinary research area where analytical methods combined with powerful stochastic optimization and metaheuristic approaches produced human-competitive results. From an engineering standpoint, ECV aims to design software and hardware solutions useful for solving challenging CV problems. From a scientific viewpoint, the goal is to enhance our current understanding of visual processing in nature and replicate this within a seeing machine. ECV is a well-established research discipline as evolutionary algorithms are more efficient than classical optimization approaches for the discontinuous, nondifferentiable, multimodal, and noisy search, optimization, and learning problems arising in many CV tasks. EC has also demonstrated its ability as a robust approach to cope with the fundamental steps of image processing, image analysis, and image understanding included in the CV pipeline (e.g., restoration, segmentation, registration, classification, reconstruction, or tracking).

This article presents an Industry 4.0 compliant ear biometric recognition technique using dense convolutional network (DenseNet), a well-known convolutional neural network model. Compared to other biometric traits, ear recognition has been a challenge due to the unavailability of a large number of images and, therefore, the improvements due to deep learning application are still unexplored. Additionally, ear biometrics has the natural advantage of privacy preservation through excellent feature encoding, which is not yet explored. In this article, the performance of DenseNet is initially tested on typically challenging benchmarks, such as street view house numbers, Canadian Institute for advanced research, and ImageNet, achieving state-of-the-art results and requiring minimal computation time and memory. All the experiments are performed on six popular ear databases namely mathematical analysis of images, annotated web ears (AWE), extended AWE (AWE-X), computer vision laboratory ear (CVLE), Indian Institute of Technology-Delhi, and West Pomeranian University of Technology, indicating that the proposed algorithm achieves a better performance over state-of-the-art. Due to less trainable parameters and fast processing, this Industry 4.0 compliant proposed recognition method can be widely used over Internet of Biometric Things, ensuring the privacy preservation.

This paper introduces the application of a new GP based Focus of Attention technique capable of improving the accuracy level when using a Laser Pointer as an interactive device. Laser Pointers have been previously employed in combination with environment control systems as interaction devices, allowing users to send orders to devices. Accurate detection of laser spots is required for sending correct orders; moreover, false offs must be eradicated, thus preventing devices to autonomously activate/deactivate when orders have not been sent by users. The idea here is to apply a self-adjusting process to a GP based algorithm capable of focusing the attention of a visual recognition system on a narrow area of an image, where laser spots will be then located. Images are taken by video cameras working on users' environment. The results show that the new approach improves significantly the accuracy level when laser spots are present -users sending orders- while maintains the extremely low values of false offs provided by previous techniques.

Nowadays, object recognition based on local invariant features is widely acknowledged as one of the best paradigms for object recognition due to its robustness for solving image matching across different views of a given scene. This paper proposes a new approach for learning invariant region descri ptor operators through genetic programming and introduces another optimization method based on a hill-climbing algorithm with multiple re-starts. The approach relies on the synthesis of mathematical expressions that extract information derived from local image patches called local features. These local features have been previously designed by human experts using traditional representations that have a clear and, preferably mathematically, well-founded definition. We propose in this paper that the mathematical principles that are used in the description of such local features could be well optimized using a genetic programming paradigm. Experimental results confirm the validity of our approach using a widely accepted testbed that is used for testing local descriptor algorithms. In addition, we compare our results not only against three state-of-the-art algorithms designed by human experts, but also, against a simpler search method for automatically generating programs such as hill-climber. Furthermore, we provide results that illustrate the performance of our improved SIFT algorithms using an object recognition application for indoor and outdoor scenarios.

The goal of purposive or behavioral vision is to study the interactions of a visual system with the real world, creating a balance between perception and action. It is said that a system that accomplishes a visuomotor task needs to implement a selective perception process allowing specific motionaction commands. This combination is understood as a visual behavior. This paper describes a real-working system, consisting of a robotic manipulator in a hand-eye configuration, which is used as a research platform in order to evolve a specialised visual routine capable of estimating specific motion-actions. The core idea is to evolve a conspicuous point detector, based on the artificial dorsal stream model, with the purpose of using this detector inside a simultaneous localization and map building system. Experimental results show as a proof-of-concept several interesting ideas; first, that it is in fact possible to find prominent points in an image through a visual attention process; and second, that the proposed system is able to design specific visual behaviors.

In this paper we present a new approach for laser-based environment device control systems by laser pointer for handicapped people. The paper proposes the design of a Fuzzy Rule Base System for laser pointer detection. The idea is to improve the success rate of the previous approaches decreasing as much as possible the false offs, i.e., the detection of a false laser spot (since this could lead to dangerous situations). To this end, Genetic Fuzzy Systems have also been employed for improving the laser spot system detection thus reducing the system false offs, that is the main objective in this problem. The system presented in this paper, using a Fuzzy Rule Base System adjusted by a Genetic Algorithm, shows a better success rate, and the most important thing, the not desired false offs are completely avoided.

Local image features can provide the basis for robust and invariant recognition of objects and scenes. Therefore, compact and distinctive representations of local shape and appearance has become invaluable in modern computer vision. In this work, we study a local descriptor based on the Hölder exponent, a measure of signal regularity. The proposal is to find an optimal number of dimensions for the descriptor using a genetic algorithm (GA). To guide the GA search, fitness is computed based on the performance of the descriptor when applied to standard region matching problems. This criterion is quantified using the F-Measure, derived from recall and precision analysis. Results show that it is possible to reduce the size of the canonical Hölder descriptor without degrading the quality of its performance. In fact, the best descriptor found through the GA search is nearly 70% smaller and achieves similar performance on standard tests.

In this work a new method to retrieve images with similar lighting conditions is presented. It is based on automatic clustering and automatic indexing. Our proposal belongs to Content Based Image Retrieval (CBIR) category. The goal is to retrieve from a database, images (by their content) with similar lighting conditions. When we look at images taken from outdoor scenes, much of the information perceived depends on the lighting conditions. The proposal combines fixed and random extracted points for feature extraction. The describing features are the mean, the standard deviation and the homogeneity (from the co-occurrence matrix) of a sub-image extracted from the three color channels: (H, S, I). A K-MEANS algorithm and a 1-NN classifier are used to build an indexed database of 300 images in order to retrieve images with similar lighting conditions applied on sky regions such as: sunny, partially cloudy and completely cloudy. One of the advantages of the proposal is that we do not need to manually label the images for their retrieval. The performance of our framework is demonstrated through several experimental results, including the improved rates for images retrieval with similar lighting conditions. A comparison with another similar work is also presented.

Nowadays, object recognition is widely studied under the paradigm of matching local features. This work describes a genetic programming methodology that synthesizes mathematical expressions that are used to improve a well known local descriptor algorithm. It follows the idea that object recognition in the cerebral cortex of primates makes use of features of intermediate complexity that are largely invariant to change in scale, location, and illumination. These local features have been previously designed by human experts using traditional representations that have a clear, preferably mathematically, well-founded definition. However, it is not clear that these same representations are implemented by the natural system with the same structure. Hence, the possibility to design novel operators through genetic programming represents an open research avenue where the combinatorial search of evolutionary algorithms can largely exceed the ability of human experts. This paper provides evidence that genetic programming is able to design new features that enhance the overall performance of the best available local descriptor. Experimental results confirm the validity of the proposed approach using a widely accept testbed and an object recognition application.

Remote sensing has become a powerful tool to derive biophysical properties of plants. One of the most popular methods for extracting vegetation information from remote sensing data is through vegetation indices. Models to predict soil erosion like the "Revised Universal Soil Loss Equation" (RUSLE) can use vegetation indices as input to measure the effects of soil cover. Several studies correlate vegetation indices with RUSLE's cover factor to get a linear mapping that describes a broad area. The results are considered as incomplete because most indices only detect healthy vegetation. The aim of this study is to devise a genetic programming approach to synthetically create vegetation indices that detect healthy, dry, and dead vegetation. In this work, the problem is posed as a search problem where the objective is to find the best indices that maximize the correlation of field data with Landsat5-TM imagery. Thus, the algorithm builds new indices by iteratively recombining primitive-operators until the best indices are found. This article outlines a GP methodology that was able to design new vegetation indices that are better correlated than traditional man-made indices. Experimental results demonstrate through a real world example using a survey at "Todos Santos" Watershed, that it is viable to design novel indices that achieve a much better performance than common indices such as NDVI, EVI, and SAVI.

No abstract available.

This paper presents how it is possible to increase the Genetic Programming (GP) Computing Power (CP) for free, via Volunteer Computing (VC), using the well known framework BOINC plus a new ``virtualization'' layer which adds all the benefits from the virtualization paradigm. Two different experiments, employing a standard GP tool and a complex GP system, are performed --with distributed PCs over several cities-- to show the free achieved CP by means of VC, without the necessity of modifying or adapting the original GP source code. The methodology can be easily extended to Evolutionary Algorithms (EAs).

An intelligent area integrates both Internet and Intranet software/hardware supports (from small embedded sensors to powerful and dynamic computing devices) to provide information about each artifact state (e.g., a power failure of the refrigerator) without user intervention. This technological sophistication can be achieved by means of service discovery systems, which allow services and users to discover, configure and communicate with other services and users. However, most of these systems only provide support for interaction between services and software clients. In order to cope with this limitation, the SEDINU system aims at supporting interactions between nomadic users and services managed by each organizational area. As users may move from one area to another to perform their tasks, this system also provides support for user-user interaction and cooperation under specific contexts (role, location and goals). This Web-based ubiquitous and cooperative working environment is enriched by a face recognition engine that allows to locate nomadic users.

This paper describes a speciation method that allows an evolutionary process to learn several robot behaviors using a single execution. Species are created in behavioral space in order to promote the discovery of different strategies that can solve the same navigation problem. Candidate neurocontrollers are grouped into species based on their corresponding behavior signature, which represents the traversed path of the robot within the environment.Behavior signatures are encoded using character strings and are compared using the string edit distance. The proposed approach is better suited for an evolutionary robotics problem than speciating in objective or topological space. Experimental comparison with the NEAT method confirms the usefulness of the proposal.

In this paper, a multiobjective (MO) learning approach to image feature extraction is described, where Pareto-optimal interest point (IP) detectors are synthesized using genetic programming (GP). IPs are image pixels that are unique, robust to changes during image acquisition, and convey highly descriptive information. Detecting such features is ubiquitous to many vision applications, e.g. object recognition, image indexing, stereo vision, and content based image retrieval. In this work, candidate IP operators are automatically synthesized by the GP process using simple image operations and arithmetic functions. Three experimental optimization criteria are considered: 1) the repeatability rate; 2) the amount of global separability between IPs; and 3) the information content captured by the set of detected IPs. The MO-GP search considers Pareto dominance relations between candidate operators, a perspective that has not been contemplated in previous research devoted to this problem. The experimental results suggest that IP detection is an illposed problem for which a single globally optimum solution does not exist. We conclude that the evolved operators outperform and dominate, in the Pareto sense, all previously man-made designs.

Recognizing and localizing objects is a classical problem in computer vision that is an important stage for many automated systems. In order to perform object recognition many researchers have focused on local features as the basis of their proposed methodologies. This work is devoted to the task of learning invariant region descriptor operators with genetic programming. The idea is to find a set of expressions that could be equal or better than the weighted gradient magnitude that is normally applied on the SIFT descriptor. This magnitude corresponds to the operator that we would like to improve through genetic programming (GP). The key for a successful problem statement was achieved with the F-measure. After a bibliographical study we have found a criterion that is simple, reliable, and useful in the estimation of such a metric. The measure that we propose here is based on the harmonic mean which is normally used by the information retrieval community. Experimental results show that the evolved descriptor¿s operator can enhance significantly the overall performance of the SIFT descriptor and surpass other state-of-the-art algorithms.

December 01 2008 Editorial Introduction to the Special Issue on Evolutionary Computer Vision In Special Collection: CogNet S. Cagnoni, S. Cagnoni Guest Editor Search for other works by this author on: This Site Google Scholar E. Lutton, E. Lutton Guest Editor Search for other works by this author on: This Site Google Scholar G. Olague G. Olague Guest Editor Search for other works by this author on: This Site Google Scholar Author and Article Information S. Cagnoni Guest Editor E. Lutton Guest Editor G. Olague Guest Editor Online Issn: 1530-9304 Print Issn: 1063-6560 © 2008 by the Massachusetts Institute of Technology2008 Evolutionary Computation (2008) 16 (4): 437–438. https://doi.org/10.1162/evco.2008.16.4.437 Cite Icon Cite Permissions Share Icon Share Facebook Twitter LinkedIn MailTo Views Icon Views Article contents Figures & tables Video Audio Supplementary Data Peer Review Search Site Citation S. Cagnoni, E. Lutton, G. Olague; Editorial Introduction to the Special Issue on Evolutionary Computer Vision. Evol Comput 2008; 16 (4): 437–438. doi: https://doi.org/10.1162/evco.2008.16.4.437 Download citation file: Ris (Zotero) Reference Manager EasyBib Bookends Mendeley Papers EndNote RefWorks BibTex toolbar search Search Dropdown Menu toolbar search search input Search input auto suggest filter your search All ContentAll JournalsEvolutionary Computation Search Advanced Search This content is only available as a PDF. © 2008 by the Massachusetts Institute of Technology2008 Article PDF first page preview Close Modal You do not currently have access to this content.

This work describes how evolutionary computation can be used to synthesize low-level image operators that detect interesting points on digital images. Interest point detection is an essential part of many modern computer vision systems that solve tasks such as object recognition, stereo correspondence, and image indexing, to name but a few. The design of the specialized operators is posed as an optimization/search problem that is solved with genetic programming (GP), a strategy still mostly unexplored by the computer vision community. The proposed approach automatically synthesizes operators that are competitive with state-of-the-art designs, taking into account an operator's geometric stability and the global separability of detected points during fitness evaluation. The GP search space is defined using simple primitive operations that are commonly found in point detectors proposed by the vision community. The experiments described in this paper extend previous results (Trujillo and Olague, 2006a,b) by presenting 15 new operators that were synthesized through the GP-based search. Some of the synthesized operators can be regarded as improved manmade designs because they employ well-known image processing techniques and achieve highly competitive performance. On the other hand, since the GP search also generates what can be considered as unconventional operators for point detection, these results provide a new perspective to feature extraction research.

This paper presents a method for object recognition, novel object detection, and estimation of the most salient object within a set. Objects are sampled using a scale invariant region detector, and each region is characterized by the subset of texture and color descriptors selected by a Genetic Algorithm (GA). Using multiple views of an object, and multiple regions per view, objects are modeled using mixtures of Gaussians, where each object represents a possible class for a particular image region. Given a set of objects, the GA learns a corresponding Gaussian Mixture Models (GMM) for each object in the set employing a one vs. all training scheme. Thence, given an input image where interest regions are detected, if a large majority of the regions are classified as regions of object O then it is assumed that said object appears in the imaged scene. The GA’s fitness function promotes: 1) a high classification accuracy, 2) the selection of a minimal subset of descriptors, and 3) a high separation among models. The separation between two GMMs is computed using a weighted version of Fisher’s linear discriminant, which is also used to estimate the most “salient” object among the set of modeled objects. Object recognition and novel object detection are done using confidence-based classification. Hence, when a non-modeled object is sampled, the detected regions are thereby identified as belonging to an unseen object and a new GMM is trained accordingly. Experimental results on the COIL-100 data set confirm the soundness of the approach.

This paper introduces a novel analogy with the way in which honeybee colonies operate in order to solve the problem of sparse and quasi dense reconstruction. To successfully solve increasingly complex problems, we must develop effective techniques for evolving cooperative solutions in the form of interacting coadapted subcomponents. A new adaptive behavior strategy is presented based on the "divide and conquer" approach used by the honeybee colony to solve search problems. The general ideas that explain the honeybee behavior are translated into a computational algorithm following the evolutionary computing paradigm. Experiments demonstrate the importance of the proposed communication system to reduce dramatically the number of outliers.

This contribution presents a novel approach for the automatic generation of a low-level feature extractor that is useful in higher-level computer vision tasks. Specifically, our work centers on the well-known computer vision problem of interest point detection. We pose interest point detection as an optimization problem, and are able to apply Genetic Programming to generate operators that exhibit human-competitive performace when compared with state-of-the-art designs. This work uses the repeatability rate that is applied as a benchmark metric in computer vision literature as part of the GP fitness function, together with a measure of the entropy related with the point distribution across the image. This two measures promote geometric stability and global separability under several types of image transformations. This paper introduces a Genetic Programming implementation that was able to discover a modified version of the DET operator [3], that shows a surprisingly high-level of performace. In this work emphasis was given to the balance between genetic programming and domain knowledge expertise to obtain results that are equal or better than human created solutions.

We present a new bio-inspired approach applied to a problem of stereo image matching. This approach is based on an artificial epidemic process, which we call the infection algorithm. The problem at hand is a basic one in computer vision for 3D scene reconstruction. It has many complex aspects and is known as an extremely difficult one. The aim is to match the contents of two images in order to obtain 3D information that allows the generation of simulated projections from a viewpoint that is different from the ones of the initial photographs. This process is known as view synthesis. The algorithm we propose exploits the image contents in order to produce only the necessary 3D depth information, while saving computational time. It is based on a set of distributed rules, which propagate like an artificial epidemic over the images. Experiments on a pair of real images are presented, and realistic reprojected images have been generated.

In this work the problem of camera placement for automated visual inspection is studied under a multi-objective framework. Reconstruction accuracy and operational costs are incorporated into our methodology as separate criteria to optimize. Our approach is based on the initial assumption of conflict among the considered objectives. Hence, the expected results are in the form of Pareto optimal compromise solutions. In order to solve our optimization problem an evolutionary based technique is implemented. Experimental results confirm the conflict among the considered objectives and offer important insights into the relationships between solution quality and process efficiency for high-accurate 3D reconstruction systems.

We present a multiobjective methodology, based on evolutionary computation, for solving the sensor planning problem for an active vision system. The application of different representation schemes, that allow to consider either fixed or variable size camera networks in a single evolutionary process, is studied. Furthermore, a novel representation of the recombination and mutation operators is brought forth. The developed methodology is incorporated into a 3D simulation environment and experimental results shown. Results validate the flexibility and effectiveness of our approach and offer new research alternatives in the field of sensor planning.

Corner measurement is of main concern within the following tasks: camera calibration, image matching, object tracking, recognition and reconstruction. This paper presents a hybrid evolutionary ridge regression approach for the problem of corner modeling. We search model parameters characterizing L-corner models by means of fitting the model to the image data. As the model fitting relies on an initial parameter estimation, we use a global approach to find the global minimum. Experimental results applied to an L-corner using several levels of noise show the advantages and disadvantages of our evolutionary algorithm compared to down-hill simplex and simulated annealing.

Automation of photogrammetric tasks by means of manipulator robots is a complex problem. It involves many planning and controlling aspects that reflect on the overall system performance in terms of precision and efficiency. This paper deals with the problem of task distribution for a multiple manipulator work cell with the goal of obtaining highly accurate object measurements. Task distribution is separated into two independent combinatorial optimization problems: activity assignment and tour planning. These problems are solved simultaneously by an optimization method based on genetic algorithms. This method implements a series of restriction-based heuristics in order to utilize a simple genetic representation similar to random keys. Experiments that validate the effectiveness of our approach are presented.

Concerns the automation of the camera network design process in order to obtain accurate 3D measurements. We restrict ourselves to the problem where the camera positions are limited only by the incidence angle constraint and it is simplified to the case where the cameras remain at a fixed distance to the set of target points to be measured. The main question addressed is where to place the cameras in order to obtain the minimal 3D error. From this question several subproblems arise: how to develop a good criterion to judge the configuration; what conditions are needed for the system to work; which are the interrelated aspects involved in the development of the system; and how to optimize the placement of the camera. From these initial questions the choice of a criterion combined with an optimization process is the key concept. The approach can be divided into two main components. Firstly, we develop an uncertainty analysis based on error propagation. This allows us to express an error criterion to be minimized. Secondly, we present an evolutionary optimization method similar to genetic algorithm, which optimizes this criterion.