Focusing on the problem that current image matching algorithm based on local feature has low correct rate of illumination change sensitive matching, an image matching algorithm with illumination robustness was proposed. Firstly, a Real-Time Contrast Preserving decolorization (RTCP) algorithm was used for grayscale image, and then a contrast stretching function was used to simulate the influence of different illumination transformation on image to extract feature points of anti-illumination transformation. Finally, a feature point descriptor was established by using local intensity order pattern. According to the Euclidean distance of local feature point descriptor of image to be matched, the Euclidean distance was determined to be a pair matching point. In open dataset, the proposed algorithm was compared with Scale Invariant Feature Transform (SIFT) algorithm, Speeded Up Robust Feature (SURF) algorithm, the "wind" (KAZE) algorithm and ORB (Oriented FAST and Rotated, BRIEF) algorithm in matching speed and accuracy. The experimental results show that with the increase of image brightness difference, SIFT algorithm, SURF algorithm, the "wind" algorithm and ORB algorithm reduce matching accuracy rapidly, and the proposed algorithm decreases matching accuracy slowly and the accuracy is higher than 80%. The proposed algorithm is slower to detect feature points and has a higher descriptor dimension, with an average time of 23.47 s. The matching speed is not as fast as the other four algorithms, but the matching quality is much better than them. The proposed algorithm can overcome the influence of illumination change on image matching.

A common spatial query often leads to the problem of multiple query results because a spatial database usually contains large size of data. To deal with this problem, a new categorization approach for spatial database query results was proposed. The solution consists of two steps. In the offline step, the coupling relationship between spatial objects was evaluated by considering the location proximity and semantic similarity between them, and then a set of clusters over the spatial objects could be generated by using probability density-based clustering method, where each cluster represented one type of user requirements. In the online query step, for a given spatial query, a category tree for the user was dynamically generated by using the modified C4.5 decision tree algorithm over the clusters, so that the user could easily select the subset of query results matching his/her needs by exploring the labels assigned on intermediate nodes of the tree. The experimental results demonstrate that the proposed spatial object clustering method can efficiently capture both the semantic and location relationships between spatial objects. The query result categorization algorithm has good effectiveness and low search cost.

Concerning the problem that how to initialize the weights of deep neural networks, which resulted in poor solutions with low generalization for spam filtering, a classification method of Deep Belief Net (DBN) was proposed based on the fact that the existing spam classifications are shallow learning methods. The DBN was pre-trained with the greedy layer-wise unsupervised algorithm, which achieved the initialization of the network. The experiments were conducted on three datesets named LinsSpam, SpamAssassin and Enron1. It is shown that compared with Support Vector Machines (SVM) which is the state-of-the-art method for spam filtering in terms of classification performance, the spam filtering using DBN is feasible, and can get better accuracy and recall.