| In the process of industrial banding bar operations, for its recognition and count has important research value. Artificial count, however, is a heavy and low efficiency,factory environment and worker individual factors will affect the accuracy of banding bar number.In order to efficiently obtain the banding number of bars, through investigations it can be efficiently used in industrial environments hardware, select Android embedded development system to ensure that the hardware can be collected to the original image quality, to ensure recognition counting algorithm can efficiently run.Identify the counting equipment uses the integrated design is more conducive to the use of industrial grade.Through many experiments and analysis, this paper presents an adaptive method for obtaining bar diameter R; according to the morphological characteristics of the banding bar proposed search algorithm and criterion for adhesive type bars;take into account the handheld device and image acquisition difficult to fix the distance between the bars, leading to the size of bar has no practical significance, it proposed a morphological parameters intelligent judgment bar diameter R, and definitions given common morphological parameters, elaborated on this entire calculation process system algorithm.Cross-section of corrosion rust and other issues for the tying bar-intensive,environmental factors, first of all, on the gray bar sectional image processing;secondly, according to the gray-scale image histogram analysis of the two peaks and one trough features, binarization processing; then, the binary image expansion,corrosion and other operations, to solve the sticking problem; finally, the processed binary image search and counted.The system realizes the OpenCV image processing development model in the Android system,designed a kind of system stability handheld hardwaredevices,developed for the banding bar count Android application software,this software function is perfect, simple operation, good interactive performance.On this basis, adding artificial correction links,after experimental verification, banding bars can be accurate to 100%.The operation speed,high accuracy and implements the algorithms of image acquisition, processing, database storage, data query results in the integration of design and development mode, meet the demand of the industrial grade. |
