Lamei Zou

ATR algorithm performance evaluation based on the simulation image and real image

Currently there is no algorithm which can be adapted to all of the imaging conditions. So, it is necessary for us to find a method to evaluate the existing ATR (automatic target recognition) algorithm. We do some researches on ATR algorithm performance evaluation based on test methodology. The basic idea of the algorithm performance evaluation is to establish the relationship model between the image quality characteristics and the algorithm’s performance. In this paper, the algorithm performance evaluation’s techniques are studied, which include the algorithm performance assessment framework, the universal test image database’s creating, and the research of the image quality evaluation model. Firstly, under the guidance of the orthogonal experimental design method, we construct a universal test image database which includes the simulation image and the outfield flight data. And then this paper propose a method to establish the relation model between image quality characteristic and ATR algorithm based on SVM classifier. Finally we use the model to evaluate algorithm’s performance. We conduct experiments on the matching algorithm’s performance evaluation. The experimental results show that the proposed evaluation framework is efficient and the evaluation model is well.

The Application of Run-Length Features in Remote Sensing Classification Combined with Neural Network and Rough Set

In this paper, we propose a method of remote sensing classification based on run-length features combined with neural network. According to the criterion of variances between & within classes, we choose efficient features and exclude redundant ones successfully with the method of rough set. In experiment, we use run-length features, co-occurrence features, gray-level gradient co-occurrence features and gray-level smoothed co-occurrence features respectively as inputs of three types of classifiers: BP net, RBF net and a nearest neighbor classifier: K-NN method when applying remote sensing classification for large scale panchromatic SPOT images with high spatial resolution. The result demonstrates the efficiency of the method proposed in this paper.

Single image super-resolution based on convolutional neural networks

We present a deep learning method for single image super-resolution (SISR). The proposed approach learns end-to-end mapping between low-resolution (LR) images and high-resolution (HR) images. The mapping is represented as a deep convolutional neural network which inputs the LR image and outputs the HR image. Our network uses 5 convolution layers, which kernels size include 5×5, 3×3 and 1×1. In our proposed network, we use residual-learning and combine different sizes of convolution kernels at the same layer. The experiment results show that our proposed method performs better than the existing methods in reconstructing quality index and human visual effects on benchmarked images.

Contour detection based on brightness and contour completion

The further research of visual processing mechanism provides a new idea for contour detection. On the primary visual cortex, the non-classical receptive field of the neurons has the orientation selectivity exerts suppression effect on the response of classical receptive field, which influences edge or line perception. Based on the suppression property of non-classical receptive field and contour completion, this paper proposed a contour detection method based on brightness and contour completion. The experiment shows that the proposed method can not only effectively eliminate clutter information, but also connect the broken contour points by taking advantage of contour completion.

The design and implement of a 3-in-1 SOC for fiber module applications

The 8-bit PIC core MCU SOC chip integrates three main functions in a single chip providing optimal solutions for various fiber module applications. The functional blocks main include a 2.5Gbps Post-Amplifier (PA), a 1.25Gbps burst or continuous mode Laser-Driver (LD), and an 8-bit PIC-core MCU. The PA receives differential signal and performs the quantization amplifications. The LD driver performs the transmit function. The on-chip MCU is the flexibility and versatility of controlling and configurations of fiber module design. The main purpose of the article is that the system digital peripherals design, such as I2C slave for host communication and its associated SFP 2-port RAM, two levels of password protection for SFP 512-byte registers, and I2C master for EEPROM interface, and the access to on-chip 2K program SRAM, and boot ROM code process.

Analysis and evaluation for the robustness of feature detection algorithm in medium-wave infrared scene

Target recognition in natural scenes generally uses the remote sensing image preparation for the matching feature template. Image match system finds correspondence between real image and remote sensing image and then output the target position parameters to the aircraft. How to quantitative analysis and evaluate the robustness of the infrared features, and to determine the available features in infrared image matching recognition algorithm is one of the keys of the ground target recognition in complex scenes. In this paper, we built a feature robustness evaluation model for typical match identifying by analyzing for the robustness of features extracted from typical terrains and targets in various condition. Combined with France SE-Workbench-IR simulation platform, we designed a special scene simulation development process, in case of lack of terrain generation module, it also can generate MWIR natural scene image. By analyzing the simulation image and real-time image in the same condition, we can gain the variation information from infrared radiation characteristic in different natural condition. Finally, we verify and assess the robustness of the matching features.

Novel image matching confidence fusion evaluation algorithm based on Support Vector Machine

Confidence evaluation is an important technique in image matching process. This paper proposes a confidence level evaluation method for image matching result based on support vector machine (SVM). We divide the matching result into two different types: the correct result and the wrong result. So we translate the match results confidence evaluation problem into the matching results classification. This paper firstly provides a method of how to prepare the character parameters which can accurately reflect the matching performance. And then the SVM based on Gaussian kernel is used as a classifier to classify the match result and discriminate the match results type. The experiments show that this method is effective. Compared with the Dempster-Shafer (D-S) evidence reasoning fusion method it has much higher accuracy.