A Survey on Recent Image Indexing and Retrieval Techniques for Low-Level Feature Extraction in CBIR Systems

Automatic Detection of Surface Defects on Citrus Fruit based on Computer Vision Techniques

“…However, a color histogram based technique reported in [5][6][7][8] one of the oldest methods, many researchers admit it as a basic method for CBIR. It has influenced many works, although, Histograms hold some major shortcomings [9][10][11]. For instance, and because Histograms are unable to spatially locate colors in images, many efforts have been achieved to enhance this approach to enable taking into account the location information of colors.…”

An overview of content-based image retrieval techniques (CBIR)

“…N .shirivastara et al [10] suggested the local structural model (LSP) that shows a large amount of pixels and the center of local disputes are considered as the threshold hence it less susceptible to noise. [6] In LBP, there must give attention to integrated patterns such as curved edges or straight edges, but some images, has a complex structure with high curved edges, corners or boundaries are crossed which makes it difficult to extract using integrated LBP [6].S.liao and et al in [12], suggest the dominant local binary patterns that uses all possible patterns rather giving consideration to mere integration patterns. Thus, complex shapes in the image are record [6].…”

“…LDP generating a pixel image using the four directions, separately labeled, and 32-bit binary sequence produced by attaching four 8-bit binary sequences. The advantage of using local patterns of high-grade, is that, it is capable of delivering a stronger and more accurate information and greater recovery rate of LBP as well [6].…”

“…The effectiveness of the CBIR system, depends on the performance of the algorithm it uses to extract image features. [6] Fig. 1: Block diagram of the CBIR system [7] III.…”

“…It is the image as a whole, while the local descriptors or low-level features represent the whole picture in the all of the area. [6] (The 1 to 7 are low-level feature extraction techniques). Finally, with regard to the recently, have used data mining techniques by researchers for image retrieval and their good results achievements, we introduced a number of these techniques as follow.…”