Research on image vectorization based on mathematical morphology

Image vectorization plays an important role in the digital image processing. Because the traditional linear vectorization methods have some shortcomings including processing data slowly, being sensitive to noises and being easy to be distorted, this paper proposes an image vectorization method based on mathematical morphology. In the paper we present an improving immune genetic algorithm based on chaos theory. The over-spread character and randomness of chaos can be used to initialize population and improve the searching speed, and the initial value sensitivity of chaos can be used to enlarge the searching space. To avoid the local optimization, the algorithm renews population and enhances the diversity of population by using density calculation of immune theory and adjusting new chaos sequence.

This paper presents a novel pupil center detection method, which is robust to the presence of eyelashes and eyelids. First, a low-intensity region is extracted, and its contour is expressed using Freeman's chain code. The five successive elements of Freeman's chain code are then replaced by their mode value. Subsequently, using the filtered Freeman's chain code, the contour graph is generated as a graph in which a vertex is an element, a directed edge is an adjacent relationship, and an edge weight is a quantization direction. An abstracted contour graph is generated from the contour graph, and the pupil center is detected based on an analysis of the abstracted contour subgraph. An evaluation experiment is conducted to evaluate the performance of the proposed approach. It is thus confirmed that the iris detection rate improved although the processing time increased as compared with those in previous studies. This means that the proposed approach can improve the convenience of iris authentication.

The C-chain code, a new three-dimensional (3-D) chain code for encoding 3-D curves, is introduced. Its quantization points are placed at the centers of grids of the side surfaces of the cube, not at its corners, to achieve longer and more varieties of the link lengths. It is shown that through estimating the frequencies of individual quantization points by their spatial configurations, the efficiency of labeling quantization points of the C-chain code can be improved by applying Huffman coding. Examples are given to illustrate this new 3-D chain code to show its major advantages over Freeman's chain code. Applications of the C-chain code to the encoding of 3-D objects are also included.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

A new approach to object recognition is proposed. The main concern is on irregular objects which are hard to recognize even for a human. The recognition is based on the contour of an object. The contour is obtained with morphological operators and described with a Freeman chain code. The chain code histogram (CCH) is calculated from the chain code of the contour of an object. For an eight-connected chain code an eight dimensional histogram, which shows the probability of each direction, is obtained. The CCH is a translation and scale invariant shape measure. The CCH gibes only an approximation of the object's shape so that similar objects can be grouped together. The discriminatory power of the CCH is demonstrated on machine-printed text and on true irregular objects. In both cases it is noted that similar objects are grouped together. The results of experiments are good. It has been shown that similar objects are grouped together with the proposed method. However, the sensitivity to small rotations limits the generality of the method.

Chain codes are the most size-efficient lossless compression methods for representing rasterised binary objects and contours. Satisfactory compression ratio, low processing cost and low storage requirements of the decoder make chain code technique interesting for storage and transmission of predefined graphical objects in embedded environments. Each element in the chain is encoded to show the relative angle difference between two adjacent pixels along the boundary of an object. The cost of binary bits representing the codes are considered to be equal. Yet, more efficient encoding is possible by considering and applying technique that treats the binary bits differently considering its requirement of storage space, energy consumption, speed of execution and etc. This paper considers cost of binary digits as unequal and proposes a new representation of the eight-direction Freeman chain code based on a variation of Huffman coding technique, which considers cost of bits as unequal. The evaluation and comparison of the cost efficiency between classical Freeman chain code and the new representation of the chain code is provided. Our experiments yield that the proposed representation of Freeman Chain code reduces overall storage/transmission cost of encoded objects considerably with compared to classical Freeman chain code.

A study of compression efficiency of 3-D chain codes to represent discrete curves is described. The 3-D Freeman chain code and the five orthogonal change chain directions (5OT) chain code are compared. The 3-D Freeman chain code consists of 26 directions, in 3-D Euclidean space, with no invariance under rotation. The 5OT chain elements represent the orthogonal direction changes of the contiguous straight-line segments of the discrete curve. This chain code only considers relative direction changes, which allows us to have a curve descriptor invariant under rotation, and mirroring curves may be obtained with ease. In the 2-D domain, Freeman chain codes are widely used to represent contour curves. Until now, the authors have had no information of implementing Freeman chain codes to compress 3-D curves. Our contribution is how to implement the Freeeman chain code in 3-D and how to compare it with the recently proposed 5OT code. Finally, to probe our results, we apply the proposed method to three different cases: arbitrary curves, cube-filling Hilbert curves, and lattice knots.

Before lung cancer classification, the preprocessing work can include lung segmentation for separating the lung area from the rest; however, there are nodule conditions called juxta-pleural and pleural-tail nodules that grow on or connect to the pleural wall. These types of nodules will get excluded during lung segmentation using standard thresholding; thus, an additional method is needed to prevent this exclusion and get the nodule included within lung parenchyma. There are two known methods for addressing the juxta-pleural nodule to be included in lung parenchyma during segmentation: the Freeman Chain Code and Morphological operation. Each technique has advantages and disadvantages; Morphological Operation excels in speed but cannot address large juxta-pleural openings. 2D Chain Code can address larger juxta-pleural but fails to correct segment separation; 3D Chain Code can address most juxta-pleural but runs slowly.

Because of the shortcomings of traditional linear vector methods, this paper proposed the image vectorization method based on mathematical morphology, including the morphological sequential homotypic skeleton abstraction method based on structure elements template and the vectorization method based on the dynamic change of pace about Freemans chain code. Experimental results show that: the method has certain feasibility and practicability.

Accurate automatic recognition of handwritten Arabic numerals has several important applications, e.g. in banking transactions, automation of postal services, and other data entry related applications. A number of modelling and machine learning techniques have been used for handwritten Arabic numerals recognition, including Neural Network, Support Vector Machine, and Hidden Markov Models. This paper proposes the use of abductive networks to the problem. We studied the performance of abductive network architecture on a dataset of 21120 samples of handwritten 0-9 digits produced by 44 writers. We developed a new feature set using histograms of contour points chain codes. Recognition rates as high as 99.03% were achieved, which surpass the performance reported in the literature for other recognition techniques on the same data set. Moreover, the technique achieves a significant reduction in the number of features required.

This paper examines several line-drawing pattern recognition methods for handwritten character recognition. They are the picture descriptive language (PDL), Berthod and Maroy (BM), extended Freeman's chain code (EFC), error transformation (ET), tree grammar (TG), and array grammar (AG) methods. A new character recognition scheme that uses improved extended octal codes as primitives is introduced. This scheme offers the advantages of handling flexible sizes, orientations, and variations, the need for fewer learning samples, and lower degree of ambiguity. Finally, the simulation of off-line character recognition by the real-time on-line counterpart is investigated.

A new coding scheme is proposed for two-tone image contours. The basic idea is to detect digital line segments on the contour and code them using fixed- or variable-length codewords; the present work deals mainly with fixed-length codewords. It is demonstrated that the conventional contour run length coding by Freeman's chain code is a special case of this scheme. The data compressibility of the scheme is studied and the test results on several contours are presented. The results show that the present scheme is superior to the conventional scheme.

This paper proposes a new algorithm of the automatic personal identification using extracted contour lines and feature points from human face profiles. As a decision function for the identification, we use a norm of a 19 dimensional feature vector, the components of which are the weighted distances between two feature points and the angles between two lines connecting three consecutive feature points. The 11 feature points are extracted from a contour line of the input profile expressed by Freeman's chain code using digital curvatures of the line. The effects of deformation of profiles caused by face panning and tilting and mouth opening upon identification accuracy have been investigated in detail. To overcome the deformation effects we propose to register three profiles per person: a normal head position profile, a tilted profile and a panned profile. The simulation results by 68 subjects show that the identification accuracy for the same persons is 91.7% and discrimination accuracy for different persons is 99.9%. This proves the superiority of the proposed algorithm.

This paper discusses a VLSI implementation of a fast contour tracing algorithm. Since Freeman published his chain code method for encoding the contour of binary images in 1961, this contour tracing method has been used in many different applications. In this paper, Freeman's chain code algorithm has been reformulated so that it can be implemented in array processing. A complete look-up table was first presented and we have derived from that an adaptive selection-inhibition circuit coupled with a priority encoder. From a VLSI implementation point of view, those circuits not only have area and speed advantages, but also are easy to lay out.

Fractal dimension is an important tool for the analysis of medical images. MR images and the mammographic images give a detailed view of both benign and malignant lesions. This work focuses on the extraction of boundary and analysis of its fractal dimensions to classify the benign and malignant lesions. The boundary of the lesion is extracted using fuzzy c-means clustering. The lesion contour is then converted into time series using “freeman’s chain code”. The resultant time series is used for fractal analysis using non-linear technique called Higuchi algorithm. This algorithm generates a single value which discriminates between benign and malignant structures. This classification of lesions is made using k-nearest neighbor pattern classifier. As a means of confirming the result the histopathological images of the tissues were used. The steps involved include selecting the region of interest of the histopatholgical images and analysis of its fractal dimensions to classify the benign and malignant lesion cells. Box-Counting algorithm is applied on the selected region to obtain various fractal features such as coarseness, lacunarity and complexity of the cells. Based on the obtained values classification of lesions is made using k-nearest neighbor pattern classifier. The results proved that these methods serve as an effective tool for diagnosis.Keywords: Box-Counting, c-means Clustering, Coarseness and Complexity, Fractal Dimension, Freeman’s Chain Code, Higuchi Algorithm, k-NN Classifier, Lacunarity, MATLAB

A new, high speed method filling in the contours of alpha-numeric characters which produces accurate and correct image patterns is proposed. The method is called the improved edge fill method because it improves on a previously developed edge fill method. It does so by selecting fill pixels from contour lines expressed by Freeman's chain code and using combinations for neighboring the codes and accordingly processing time is reduced by about 20% over conventional methods. For binary image contour filling, the method eliminates faults found in conventional edge fill methods, and in consequence the areas inside contour lines are filled in rapidly and correctly. The method's effectiveness is examined in experiments using both Arabic numerals and letters of the Roman alphabet. As a result it is found that this filling method is able to produce correct and accurate image patterns and that it can be applied to alpha-numeric character contour filling. >