Wenzhu Yang

An improved genetic algorithm for optimal feature subset selection from multi-character feature set

This paper presents an improved genetic algorithm (IGA) by which the optimal feature subset can be selected effectively and efficiently from a multi-character feature set (MCFS). IGA adopts segmented chromosome management scheme to implement local management of chromosome. This scheme encodes a solution with an entire binary chromosome; but logically, it divides the chromosome into several segments according to the number of feature groups in MCFS for local management. A segmented crossover operator and a segmented mutation operator are employed to operate on these segments to avoid invalid chromosomes. The probability of crossover and mutation are adjusted dynamically according to the generation number and the fitness value. As a result, IGA obtains strong searching ability at the beginning of the evolution and achieves accelerated convergence along the evolution. IGA is tested using features extracted from cotton foreign fiber objects, and compared with the Simple Genetic Algorithm (SGA) under the same condition. The results show that IGA receives improved searching ability and convergence speed compared with SGA. The optimal feature subset selected by the IGA has much smaller size than that of the SGA. This is very important for the online classification of foreign fibers.

An Automated Visual Inspection System for Foreign Fiber Detection in Lint

This paper presents an automated visual inspection system for the detection of foreign fibers in lint. The system mainly includes six parts, namely Input Conveyer, Lint Layer Generator, Output Roller, Image Acquisition System, Host Computer and Lint Layer Collector. The lint for inspection is fed into the Lint layer Generator through the Input Conveyer, and then made into uniform thin layer. The generated lint layer are dragged out of the Generator and transferred to the inspection box where two cameras are served for imaging. Live images of lint with foreign fibers are captured by the Image Acquisition System and then processed in the Host Computer using inspection algorithm designed for detection of foreign fibers. The inspection algorithm contains four main steps, namely, image enhancement, segmentation, post-processing and decision-making. Dozens of carefully selected foreign fiber samples were used to test the performance of the automated visual inspection system. The results indicate that the proposed system can detect out most of the foreign fibers mixed in lint correctly.

Feature Selection for Cotton Foreign Fiber Objects Based on PSO Algorithm

Due to large amount of calculation and slow speed of the feature selection for cotton fiber, a fast feature selection algorithm based on PSO was developed. It is searched by particle swarm optimization algorithm. Though search features by using PSO, it is reduced the number of classifier training and reduced the computational complexity. Experimental results indicate that, in the case of no loss of the classification performances, the method accelerates feature selection.