Abstract- The ancient pottery is often found in archaeological sites in a broken state, especially when those pieces of unknown organisms and irregular fragments, may take years of hard work, especially in the case of loss of some pieces or require hard work and experienced archaeologists. So this problem is divided into two major tasks the first of which is the Classification of Archaeological Fragments into similar groups (CAF) and the second one is the Reconstruction of each group into the original Archaeological Objects (RAO). To solve this problem, a method has been proposed, which exploits the color and texture properties of the surfaces of the fragments. Furthermore, the reconstruction of archaeological fragments in 3D geometry is an important problem in pattern recognition. Therefore, this research has implemented the algorithms to reconstruct real datasets using Neural Networks. The challenge of this work is to reconstruct the objects without previous knowledge about the part that should start the assembly; this greatly helps to avoid the presence of gaps created due to missing artifact fragments. The study utilizes the geometric features of the fragments as important features to reconstruct the objects by classifying their fragments using a Neural Network model.

Introduction - An automatic reconstruction of ancient artifact fragments is a great interest in archaeology. It is considered important because it helps archaeologists access inferences about past cultures and civilizations Hristov and Agre (2013). Although variety algorithms have been proposed to reconstruct archaeological pottery fragments, few studies approached the classification of the fragments found in archaeological sites into similar groups Makridis and Daras (2012). In order to highlight the most important methods that the authors adopted over the past few decades for the classification of archaeological fragments in light of extracted features, most of the previous research was shown to rely on color feature Kampel and Sablatnig (2000), texture features Ying and Gang (2010), color and texture features Smith et al. (2010), Zhou et al. (2011), color and edges Makridis and Daras (2012), at last the profile, edge, color and texture Piccoli et al. (2013).
One of the main challenges is reconstructing the archaeological objects from a large number of fragments that are found in excavation sites Guoguang et al. (2016), and determining the correct match between them Guoguang et al. (2018). Occasionally, archaeological workers suffer when trying to match object fragments together, especially in the case of a presence of significant gaps in the fragments. Thus, numerous studies proposed methods for the purpose of reaching a suitable solution to reconstruct the archaeological 3D objects and returning them to their original forms such as Belenguer and Vidal (2012). Therefore, the main objectives of this work are listed as follows:
a) To propose a novel algorithm for classifying fragments depending on global and local features, specifically color and texture features, this be performed through proposed method that includes the intersection of color points of the images, and using the local binary patterns (LBP) feature that has proven to be more flexible with color feature, but requires more complex calculations.
b) To design a robust prototype for the reconstruction of 3D objects, despite the existence of the gaps, by exploiting the geometric features (especially the slope of the edges of the fragments); as well as finding the appropriate location for matching.
This paper presents a proposed framework to solve the problem of classification and reconstruction of archaeological fragments. The proposed methodology consists of two phases; each one performs a specific job, as shown in Fig. 1.

2. Literature Review
The idea of finding the possible solutions to the resemble of objects began as early as 1970, when Smith and Kristof (1970) were interested to reassemble the Egyptian Temple with computer assistance. Many studies have focused on the issue of archeology to find a solution based on two-dimensional images and three-dimensional model, whereas some of the researchers were interested in the proposed methods of classifying fragments into groups and reconstructing the archaeological objects. Smith et al. (2010) focused on the classification of two-dimensional fragments based on the properties of color and texture. While other authors deal with the problem on the basis of surface texture properties Ying and Gang (2010). Another type of studies (Makridis and Daras 2012) had depended on the classification of the parts by using the technique the front and rear characteristics of the pottery to improve the classification accuracy and extract features based on color information and local texture. The study of Leitao and Stolfi (2005) focused on contour information for the reconstruction of ceramic fragments. Oxholm and Nishino (2011) reassemble thin artifacts of geometrically unknown through the photometric properties of the boundary contour.
Subsequently, most previous works focused on finding pairwise matches between adjacent fragments by using color surface which is one of the traditional features, that is why the work of authors Toler-Franklin et al. (2010) were so different from the others where have relied on a multiple-features that extracted from fragments based on color, shape and normal maps. Another work was suggested by Kimia and Aras (2010), which includes a framework or a practical system can be used by archaeologists to assembling 2D vessel fragment archaeological and that could be applied on the 3D fragments. By using the morphology profile, the authors Karasik and Smilansky (2011) were proposed a method that relies on the computerized morphological classification of ceramics. Oxholm and Nishino (2013) didn t adopt the shape of the object or its painted texture, but their work depended on similar geometry and photometry along, and across matching fragments adjoining regions. In the three-dimensional model case, much work has been done on the problem of automatically reconstructing fragmented objects. The authors Lu et al. (2007) provided an approach to reconstruct the fragments depended on boundary curves of the fragments and the interaction with archaeologists. Through a collaborative project Cohen et al. 2010 which considered as a formed a generic model based on the combined between the expert feedback to the archaeologist and vessel surface markings.
This model has been tested by using a ceramic artifact collection recovered from the National Constitution Center site in Independence National Historical Park. A method that classifies and reassembly of archaeological fragments based on the discriminating feature descriptors was proposed by Guoguang et al. (2016a), whereas Angelo et al. (2018) focused on analyses of pottery fragments by extracting 3D geometrical and morphological features. Then Precision and durability have emerged in the work of continuous fragments in the dimensional analysis of certain recognized properties.