Grouping genetic algorithm based image attribute clustering method

Abstract

The invention provides a grouping genetic algorithm based image attribute clustering method. The image attribution clustering method comprises the following steps of 1, randomly generating N chromosomes as an initial group with regard to N images, wherein each chromosome represents one probable attribute clustering result, each image has m attributes, and the attributes of the images are classified to a plurality of groups; 2, calculating a fitness function value of each chromosome; 3, selecting the N chromosomes of a next generation by using a roulette wheel selection strategy according to the calculated fitness function value of the chromosome; 4, executing crossover operator on a group part of the chromosomes; 5, executing mutation operator on an attribute part of the chromosomes; 6, executing inversion operator on the attribute part of the chromosomes; 7, repeating the step 2 to the step 6 until the given genetic algebra is completed; and 8, outputting the given genetic algebra which is completed to obtain the chromosome with the optimal fitness function value.

Description

technical field [0001] The present invention relates to the field of image recognition, more specifically, the present invention relates to an image attribute clustering method based on group genetic algorithm. Background technique [0002] Learning-based image recognition methods have made great progress in the past few years. For specific object classes, especially faces and cars, there are relatively reliable and efficient recognition based on underlying features (such as SIFT (ScaleInvariantFeatureTransform) features or HOG (HistogramofOrientedGradient) features). But the underlying features of these images cannot reflect the image category information well. Recently, new research materials propose methods for classification using intrinsic properties of images. Attributes refer to the characteristics that can be named by people and can be observed in the image. It can indicate whether the object in the image exists, can describe the color, shape, material, component, ...

AI Technical Summary

Problems solved by technology

Genetic algorithm is good at global search, but its local search ability is obviously insufficient, the speed of searching for the optimal solution or satisfactory solution is too slow, and there are problems such as easy to fall into the local optimal solution and premature convergence caused by excessive selection pressure.

Moreover, due to operations such as cross-mutation, different chromosomes may represent the same attribute clustering results, which increases the scope of the search and wastes time.

Images