Due to the large number and small size of electronic components, the identification speed is slow and the accuracy is low. This research develops an electronic component identification system based on FPGA and machine learning technology. The core of the system is FPGA, combined with an efficient image processing unit to achieve parallel acceleration of machine learning model calculations. To extract features, we use Hu matrix methods that are robust to position translation, size scaling, and image rotation, which are then used as input data for the convolutional neural network. The network consists of 8 layers and uses ReLU activation function. Experiments show that the system shows a high degree of accuracy and processing speed in different types of part identification, with an average accuracy of 99.15% and an average time of 551.5 milliseconds. The use of FPGA resources and experimental results verify the effectiveness and superior performance of the system. The system has wide application potential in the field of industrial automation, and has certain popularization value in providing efficient and accurate identification of electronic components.