Search

Search Results

• DESIGNING MLP AND CNN NEURAL NETWORK MODULES ON FPGA FOR IMAGE CLASSIFICATION TASKS

  E. V. Melnik , D.Е. Blokh , А.I. Bezmeltsev , V.S. Panishchev , S.N. Poltoratsky

  214-229

  2025-11-10

  Abstract ▼

  Relevance. The development of machine learning methods and neural network architectures, as well as their spread into various industrial sectors, determine the relevance of solving problems related to their hardware implementation. The use of programmable logic integrated circuits in this area will increase data processing speed and the adaptability of the implemented algorithms. However, designing neural network architectures on programmable logic integrated circuits is associated with a number of methodological and technical difficulties, including the optimization of parallel computing, hardware resource management, and ensuring operation under conditions of limited computing resources. The purpose of this work is to analyze and compare two neural network architectures, the multilayer perceptron (MLP) and the convolutional neural network (CNN), in the context of their hardware implementation on programmable logic integrated circuits (PLICs). Particular attention is paid to the trade-off between classification accuracy and the efficient use of limited FPGA hardware resources. Research methods.
  To achieve the goal, two modules were developed and simulated on a Virtex 7 FPGA, a perceptron and a convolutional module. The MNIST dataset, reduced to 20×20 pixels, was used. The implementation included quantizing parameters to a fixed 16:16 format, optimizing hyperparameters, using tabular computations for nonlinear functions, and evaluating FPGA resource usage. Results and discussions.
  MLP achieved 93% accuracy using 11% of logic elements, while CNN achieved 98% accuracy but required significantly more resources. The use of internal buffers to store intermediate data in CNN resulted in exceeding the allowable resources. The forced transition to external memory increased delays and the number of I/O ports. Conclusions. The study showed that the choice of architecture depends on priorities: CNN provides better accuracy but is less resource-efficient. For embedded systems with memory and power consumption constraints, a simplified MLP implementation is preferable. The main problems remain the lack of internal memory and the high resource intensity of operations, which requires further research in the field of hardware optimization and adaptive computation control

• METHOD FOR RECOGNIZING TEXT DATA IN IMAGES

  V.S. Panishchev, О.О. Khomyakov, D.V. Titov, S.I. Egorov

  2023-10-23

  Abstract ▼

  The purpose of the study is to study the problems arising in the process of digital image
  processing in systems for obtaining textual characteristics of product objects. Such as the selection
  of objects that fall into the frames of the video stream and the recognition of text markings,
  without the use of specialized hardware. In particular, the problems that arise when working
  with images containing different levels of noise and distortion. The objectives of the study i nclude
  a comparative and analytical examination of diverse methods and algorithms utilized in
  the realm of digital image processing. The primary objectives include identifying, segmenting,
  and classifying text-containing portions within the video stream. The study aims to construct a
  mathematical model for text extraction from video frames that is adaptable to a wide array of
  objects. Evaluate recognition accuracy under varying levels of noise and perform a comparative
  analysis against alternative solutions based on the acquired data. The presented system
  analyzes the frames of the video stream and classifies the characteristics of the products in the
  frame. The solution demonstrates the behaviour and capabilities of digital image processing
  methods in various conditions in relation to the tasks of text classification and object search in
  a video stream. During the development of this system, a comparison of various options for recognizing
  symbolic information was carried out.