For the intelligent driving field, it is necessary to obtain high-quality super-resolution images under the condition of limited memory. Therefore, a vehicle-based image super-resolution reconstruction algorithm based on weighted eight-bit binary quantization was proposed. Firstly, the information compression module was designed based on the eight-bit binary quantization convolution, reducing the internal redundancy, enhancing the information flow in the network, and improving the reconstruction rate. Then, the whole network was composed of a feature extraction module, a plurality of stacked information compression modules and an image reconstruction module, and the information of the interpolated super-resolution space was fused with the image reconstructed by the low-resolution space, improving the network expression ability without increasing the complexity of the model. Finally, the entire network structure in the algorithm was trained based on the Generative Adversarial Network (GAN) framework, making the image have better subjective visual effect. The experimental results show that, the Peak Signal-to-Noise Ratio (PSNR) of the proposed algorithm for the reconstructed vehicle-based image is 0.22 dB higher than that of Super-Resolution using GAN (SRGAN), its generated model size is reduced to 39% of that of the Laplacian pyramid Networks for fast and accurate Super-Resolution (LapSRN), and the reconstruction speed is improved to 7.57 times of that of LapSRN.

The traditional sparse coding image classification algorithms extract single type features, ignore the spatial structure information of the images, and can not make full use of the feature topological structure information in feature coding. In order to solve the problems, a image classification algorithm based on multi-scale feature fusion and Hessian Sparse Coding (HSC) was proposed. Firstly, the image was divided into sub-regions with multi-scale spatial pyramid. Secondly, the Histogram of Oriented Gradient (HOG) and Scale-Invariant Feature Transform (SIFT) were effectively merged in each subspace layer. Then, in order to make full use of the feature topology information, the second order Hessian energy function was introduced to the traditional sparse coding target function as a regularization term. Finally, Support Vector Machine (SVM) was used to classify the images. The experimental results on dataset Scene15 show that, the accuracy of HSC is 3-5 percentage points higher than that of Locality-constrained Linear Coding (LLC), while it is 1-3 percentage points higher than that of Support Discrimination Dictionary Learning (SDDL) and other comparative methods. Time-consuming experimental results on dataset Caltech101 show that, the time-consuming of HSC is about 40% less than that of the Multiple Kernel Learning Sparse Coding (MKLSC). The proposed HSC can effectively improve the accuracy of image classification, and its efficiency is also better than the contrast algorithms.