Convolutional Neural Networks (CNNs) based algorithms have been successful in solving image recognition problems, showing very large accuracy improvement. In recent years, deconvolution layers are widely used as key components in the state-of-the-art CNNs for end-to-end training and models to support tasks such as image segmentation and super resolution. However, the deconvolution algorithms are computationally intensive which limits their applicability to real time applications. Particularly, there has been little research on the efficient implementations of deconvolution algorithms on FPGA platforms which have been widely used to accelerate CNN algorithms by practitioners and researchers due to their high performance and power efficiency. In this work, we propose and develop deconvolution architecture for efficient FPGA implementation. FPGA-based accelerators are proposed for both deconvolution and CNN algorithms. Besides, memory sharing between the computation modules is proposed for the FPGA-based CNN accelerator as well as for other optimization techniques. A non-linear optimization model based on the performance model is introduced to efficiently explore the design space in order to achieve optimal processing speed of the system and improve power efficiency. Furthermore, a hardware mapping framework is developed to automatically generate the low-latency hardware design for any given CNN model on the target device. Finally, we implement our designs on Xilinx Zynq ZC706 board and the deconvolution accelerator achieves a performance of 90.1 GOPS under 200MHz working frequency and a performance density of 0.10 GOPS/DSP using 32-bit quantization, which significantly outperforms previous designs on FPGAs. A real-time application of scene segmentation on Cityscapes Dataset is used to evaluate our CNN accelerator on Zynq ZC706 board, and the system achieves a performance of 107 GOPS and 0.12 GOPS/DSP using 16-bit quantization, and supports up to 17 frames per second for 512x512 image inputs with a power consumption of only 9.6W.

No abstract

No abstract

Purpose: The Yellow River delta boasts rich land resources but lacks fresh water and exhibits poor natural conditions. To rationally develop and utilize the land resources therein, it is necessary to evaluate the soil quality. Methods: Adopting specific screening conditions, principal component analysis (PCA) was used to construct a minimum data set (MDS) from 10 soil indicators. Then, a complete soil quality evaluation index system of the Yellow River delta was developed. The soil quality comprehensive index (SQI) method was used to assess the soil quality in the Kenli District, and the soil quality grades and spatial distribution were analyzed. Results: (1) The average SQI of the Kenli District is 0.523, and the best soil quality is concentrated near the Yellow River, especially in Huanghekou town. (2) The normalized difference vegetation index was positively correlate with SQI, whereas Dr (nearest distance between the sampling site and Yellow River) and Ds (nearest distance between the sampling site and Bohai Sea) were negatively correlated with SQI. Elev (sampling site elevation) was not correlated with SQI. (3) The SQI of agricultural planting is greater than that of the natural land type and significantly greater than that of nudation. The main factors limiting farmland soil quality are SK (water-soluble potassium) and pH, whereas the factor limiting the natural land type are the soil nutrient indicators. Conclusions: To improve soil quality and develop and utilize land resources, the towns should adopt systematic land development/utilization methods based on local conditions. These results have important guiding significance and practical value for the more objective and accurate evaluation of soil quality in coastal areas and the development and utilization of land resources.