【摘要】 车间工件如齿轮、轴承等是车间物料识别的基本对象,为提高对这些车间工件的识别效率和识别速度,文中提出了采用卷积神经网络结构框架进行视觉识别。根据工件的位置、尺寸、结构等特征,构造卷积神经网络识别模型,预处理采集到的图片信息,提取图像的目标区域,对车间工件进行实时检测识别,软件的运行结果证明了人工神经网络模型检测识别能力的准确性和效率。更多还原

【Abstract】 Since workshop workpieces such as gears and bearings are the basic objects of workshop material recognition, to improve the recognition efficiency and speed, the idea of applying convolution neural network structure framework to visual recognition is put forward. According to the position, size, structure and other characteristics of workpieces, a recognition model of convolutional neural network was established. By preprocessing the collected image information, target areas of images were extracted, so that the workshop workpiece can be detected and recognized in real time. Running results of the software prove the accuracy and efficiency of using artificial neural network model for recognition.更多还原