CoChat

FPGA Implementation for CNN-Based Optical Remote Sensing Object Detection

Open access
Zhang, Ning, Wei, Xin, Chen, He, Liu, Wenchao · 2021 · Electronics

The DOI matches its Crossref record; 49 of the 49 references it cites were found in a registry or an index.

Abstract

In recent years, convolutional neural network (CNN)-based methods have been widely used for optical remote sensing object detection and have shown excellent performance. Some aerospace systems, such as satellites or aircrafts, need to adopt these methods to observe objects on the ground. Due to the limited budget of the logical resources and power consumption in these systems, an embedded device is a good choice to implement the CNN-based methods. However, it is still a challenge to strike a balance between performance and power consumption. In this paper, we propose an efficient hardware-implementation method for optical remote sensing object detection. Firstly, we optimize the CNN-based model for hardware implementation, which establishes a foundation for efficiently mapping the network on a field-programmable gate array (FPGA). In addition, we propose a hardware architecture for the CNN-based remote sensing object detection model. In this architecture, a general processing engine (PE) is proposed to implement multiple types of convolutions in the network using the uniform module. An efficient data storage and access scheme is also proposed, and it achieves low-latency calculations and a high memory bandwidth utilization rate. Finally, we deployed the improved YOLOv2 network on a Xilinx ZYNQ xc7z035 FPGA to evaluate the performance of our design. The experimental results show that the performance of our implementation on an FPGA is only 0.18% lower than that on a graphics processing unit (GPU) in mean average precision (mAP). Under a 200 MHz working frequency, our design achieves a throughput of 111.5 giga-operations per second (GOP/s) with a 5.96 W on-chip power consumption. Comparison with the related works demonstrates that the proposed design has obvious advantages in terms of energy efficiency and that it is suitable for deployment on embedded devices.

Overview

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.3390/electronics10030282

Cites

49 of 49 references verified0 citation errors · 0 not checkable

Statistics

Figures

Figure 1 The structure of the improved YOLOv2. Each convolutional layer contains not only convolution but also batch normalization and activation sublayers. Figure 1. The structure of the improved YOLOv2. Each convolutional layer contains not only convolution but also batch normalization and activation sublayers.
Figure 1 The structure of the improved YOLOv2. Each convolutional layer contains not only convolution but also batch normalization and activation sublayers. Figure 1. The structure of the improved YOLOv2. Each convolutional layer contains not only convolution but also batch normalization and activation sublayers.CC-BY · From the PDF
Figure 2 The computation of the dilated convolution with a 3 × 3 kernel and a dilation rate of 2. Figure 2. The computation of the dilated convolution with a 3 × 3 kernel and a dilation rate of 2. 2.1.3. Transposed Convolution
Figure 2 The computation of the dilated convolution with a 3 × 3 kernel and a dilation rate of 2. Figure 2. The computation of the dilated convolution with a 3 × 3 kernel and a dilation rate of 2. 2.1.3. Transposed ConvolutionCC-BY · From the PDF
Image 3
CC-BY · From the PDF
Figure 4 The forward inference calculation process of the fundamental network and an overview of our optimization for implementation. Figure 4. The forward inference calculation process of the fundamental network and an overview of our optimization for implementation.
Figure 4 The forward inference calculation process of the fundamental network and an overview of our optimization for implementation. Figure 4. The forward inference calculation process of the fundamental network and an overview of our optimization for implementation.CC-BY · From the PDF
Figure 5 The forward inference calculation process after network quantization and layer Figure 5. The forward inference calculation process after network quantization and layer fusion.
Figure 5 The forward inference calculation process after network quantization and layer Figure 5. The forward inference calculation process after network quantization and layer fusion.CC-BY · From the PDF
Figure 6 The computing pattern of the standard convolutions.
Figure 6 The computing pattern of the standard convolutions.CC-BY · From the PDF
Figure 7 ( a ) The schematic diagram of the modified 1 × 1 convolution operation process. ( b )The interpolation processing of the input feature map. If the row contains feature values, we will read the feature values and zero at intervals according to the interpolation rules; otherwise, the readout values are all equal to zero. ( c ) The modified computational process of the dilated convolution with a 1-pixel stride. ( d ) The modified computational process of the dilated convolution with a 2-pixel stride. Figure 7. ( a ) The schematic diagram of the modified 1 × 1 convolution operation process. ( b )The interpolation processing of the input feature map. If the row contains feature values, we will read the feature values and zero at intervals according to the interpolation rules; otherwise, the readout values are all equal to zero. ( c ) The modified computational process of the dilated convolution with a 1-pixel stride. ( d ) The modified computational process of the dilated convolution with a 2-pixel stride.
Figure 7 ( a ) The schematic diagram of the modified 1 × 1 convolution operation process. ( b )The interpolation processing of the input feature map. If the row contains feature values, we will read the feature values and zero at intervals according to the interpolation rules; otherwise, the readout values are all equal to zero. ( c ) The modified computational process of the dilated convolution with a 1-pixel stride. ( d ) The modified computational process of the dilated convolution with a 2-pixel stride. Figure 7. ( a ) The schematic diagram of the modified 1 × 1 convolution operation process. ( b )The interpolation processing of the input feature map. If the row contains feature values, we will read the feature values and zero at intervals according to the interpolation rules; otherwise, the readout values are all equal to zero. ( c ) The modified computational process of the dilated convolution with a 1-pixel stride. ( d ) The modified computational process of the dilated convolution with a 2-pixel stride.CC-BY · From the PDF
Figure 8 The overall diagram of the proposed hardware architecture. Figure 8. The overall diagram of the proposed hardware architecture.
Figure 8 The overall diagram of the proposed hardware architecture. Figure 8. The overall diagram of the proposed hardware architecture.CC-BY · From the PDF
Image 9
CC-BY · From the PDF
Image 10
CC-BY · From the PDF
Figure 11 The diagram of the parameter storage module, which is composed of a Double Data Rate (DDR) controller and parameters buffer. Figure 11. The diagram of the parameter storage module, which is composed of a Double Data Rate (DDR) controller and parameters buffer.
Figure 11 The diagram of the parameter storage module, which is composed of a Double Data Rate (DDR) controller and parameters buffer. Figure 11. The diagram of the parameter storage module, which is composed of a Double Data Rate (DDR) controller and parameters buffer.CC-BY · From the PDF
Figure 12 The diagram of the storage strategy in the processing system (PS)-DDR. Figure 12. The diagram of the storage strategy in the processing system (PS)-DDR.
Figure 12 The diagram of the storage strategy in the processing system (PS)-DDR. Figure 12. The diagram of the storage strategy in the processing system (PS)-DDR.CC-BY · From the PDF
Image 13
CC-BY · From the PDF
Image 14
CC-BY · From the PDF

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.

A few free checks a day · papers already verified open instantly · results are public

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

The self-check needs the paper’s full text.

Citations

Cites

Show all 49 references

Cited by

See all 85

Metadata

Authors
Zhang, N., Wei, X., Chen, H., & Liu, W.
Affiliations
  • Beijing Institute of Technology (CN)
  • Tsinghua University (CN)
Corresponding author
Liu, Wenchao
Venue
Electronics
Publisher
MDPI AG
Published
2021-01-25
Volume
10
Issue
3
Pages
282
ISSN
2079-9292 (ISSN-L)
OpenAlex
W3121787284
Cited by
85
References
46
Citations by year
2026: 5 · 2025: 21 · 2024: 15 · 2023: 18 · 2022: 21 · 2021: 5
FWCI
5.31
Type
journal article
Language
en
Open access
Locations
  • Electronics (publishedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Funders
  • National Key R & D Program of China (2017YFB0502800, 2017YFB0502804)
  • the MYHT Program of China (B0201)
Topics
  • Physical Sciences › Computer Science › Computer Vision and Pattern Recognition › Advanced Neural Network Applications
  • Physical Sciences › Computer Science › Hardware and Architecture › Embedded Systems Design Techniques
  • Physical Sciences › Engineering › Electrical and Electronic Engineering › Embedded Systems and FPGA Applications
Keywords
FPGA implementation, convolutional neural networks, optical remote sensing, object detection, YOLOv2, hardware architecture, energy efficiency, embedded systems, Xilinx Zynq

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: FPGA Implementation for CNN-Based Optical Remote Sensing Object Detection. https://app.cochat.ai/papers/10.3390/electronics10030282
The paper
Zhang, N., Wei, X., Chen, H., & Liu, W. (2021). FPGA Implementation for CNN-Based Optical Remote Sensing Object Detection. Electronics. https://doi.org/10.3390/electronics10030282

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.