As the resolution of images processed in real-time continues to increase, it is necessary to compress the transmitted image data, and then transmit it to the terminal to decompress and restore the image. JPEG-LS is an algorithm that supports lossless and near-lossless compression. However, the decompression of JPEG-LS images is mostly implemented in a software environment. When decompressing multiple high-resolution images, the problems of decoding speed and resource consumption are more prominent. Therefore, the implementation of the JPEG-LS image decompression algorithm on FPGA is proposed in this paper, which divides the image into blocks and adopts a parallel processing structure. After the experiment on the hardware decoder, for the 1024*2048 test image, the designed hardware decoder can improve the original software decoding time by approximately 55%.
In recent years, Siamese network algorithms based on deep learning classes have achieved better tracking accuracy and speed and become one of the research hotspots in the field of target tracking. However, the traditional Siamese network algorithm lacks a holistic view of the target and extracts shallow features, making it easy to lose track of the target in complex environments. The paper proposes a Contextual transformer network for visual recognition (CotNet) target tracking algorithm based on attentional mechanisms and contextual awareness to address this. The paper innovatively uses the CotNet50 network as the backbone network and adopts a residual network variant design scheme with a self-attention mechanism, which can enhance the feature representation capability of the network model and improve the performance of the algorithm. In addition, to handle changes in appearance during target tracking, an efficient channel attention module, and a global contextual feature module are embedded in the backbone network branch to enhance the network's overall perception of the target and improve the algorithm's tracking accuracy. The experimental results of this paper's algorithm on the VOT2018 data show that the accuracy, robustness, and EAO (Expected Average Overlap) are improved by 7.3%, 13.95%, and 11.9% respectively compared to SiamFC. It has good tracking results when dealing with complex scenes on the OTB100 dataset.
The process of combining features from two images of different sources to generate a new image is called image fusion. In order to adapt to different application scenarios, deep learning was widely used. However, existing fusion networks focued on the extraction of local information, neglected the long-term dependencies. In order to improve the defect, a fusion network based on Transformer was proposed. To accommodate our experimental equipment, we made some modifications to Transformer. A dual-branch autoencoder network was designed with detail and semantic branches, the fusion layer consists of CNN and Transformer, and the decoder reconstructs the features to get the fused image. A new loss function was proposed to train the network. Based on the results, an infrared feature compensation network was designed to enhance the fusion effect. In several metrics that we focus on, we compared with several other algorithms. As the experiments on some datasets, our method had improvement on SCD, SSIM and MS-SSIM metrics, and was basically equal to other algorithms on saliency-based structural similarity, weighted quality assessment, and dge-based structural similarity. From the experimental results, we can see that our method was feasible.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.