Scene text detection attracts much attention in computer vision, because it can be widely used in many applications such as real-time text translation, automatic information entry, blind person assistance, robot sensing and so on. Though many methods have been proposed for horizontal and oriented texts, detecting irregular shape texts such as curved texts is still a challenging problem. To solve the problem, we propose a robust scene text detection method with adaptive text region representation. Given an input image, a text region proposal network is first used for extracting text proposals. Then, these proposals are verified and refined with a refinement network. Here, recurrent neural network based adaptive text region representation is proposed for text region refinement, where a pair of boundary points are predicted each time step until no new points are found. In this way, text regions of arbitrary shapes are detected and represented with adaptive number of boundary points. This gives more accurate description of text regions. Experimental results on five benchmarks, namely, CTW1500, TotalText, ICDAR2013, ICDAR2015 and MSRA-TD500, show that the proposed method achieves state-of-the-art in scene text detection.

Text detection and recognition in a natural environment are key components of many applications, ranging from business card digitization to shop indexation in a street. This competition aims at assessing the ability of state-of-the-art methods to detect Multi-Lingual Text (MLT) in scene images, such as in contents gathered from the Internet media and in modern cities where multiple cultures live and communicate together. This competition is an extension of the Robust Reading Competition (RRC) which has been held since 2003 both in ICDAR and in an online context. The proposed competition is presented as a new challenge of the RRC. The dataset built for this challenge largely extends the previous RRC editions in many aspects: the multi-lingual text, the size of the dataset, the multi-oriented text, the wide variety of scenes. The dataset is comprised of 18,000 images which contain text belonging to 9 languages. The challenge is comprised of three tasks related to text detection and script classification. We have received a total of 16 participations from the research and industrial communities. This paper presents the dataset, the tasks and the findings of this RRC-MLT challenge.

We explored the problems which are soon to be faced while parking autonomous cars in parking lots. Like where is the closest parking slot available to the autonomous car? How to navigate to that location? What kind of parking structures could be good for autonomous cars? We also provide an initial solution which uses a central server and the graph of the parking lot to guide the cars to the closest parking slots. With experiments, we have shown that our proposed method should be effective for the controlled parking for self-driving cars.

In this letter, a motion adaptive deinterlacing algorithm based on modular neural networks is proposed. The proposed method uses different neural networks based on the amount of motion. Modular neural networks were selectively used depending on the differences between the adjacent fields. We also used motion vectors to select optimal input pixels from the adjacent fields. Motion estimation was used to find input blocks for the neural networks with minimum errors. Intra/inter-mode switching was employed to address inaccurate motion estimation problems.

Generally, deinterlacing algorithms can be either classified as intra methods or inter methods. Intra methods interpolate missing lines by using surrounding pixels in the current field. Inter methods interpolate missing lines by using pixels and the motion information of multiple fields. Neural network deinterlacing that uses multiple fields has been proposed. It provides improved performance compared to existing neural network deinterlacing algorithms that use a single field. However, when adjacent fields are very different, neural network deinterlacing that uses multiple fields may not provide good performance. To address this problem, we propose using field-MSE values as additional inputs. These MSE values can provide helpful information so that the networks can consider field differences in using multiple fields. Experimental results show that the use of the proposed algorithm results in improved performance.

In this paper, we propose to use the principal component analysis for the compression of hyperspectral images. When hyperspectral images are compressed using conventional image compression algorithms, discriminant features of original data may be lost during compression process. In order to preserve such discriminant information, we first apply a linear feature extraction method to the original data. Then, we emphasize discriminant features and use the principal component analysis in order to compress the images whose discriminant features are enhanced. Experiments show that the proposed method provides improved classification accuracies than existing compression algorithms.

In this paper, we propose compression algorithms for hyperspectral images based on the wavelet transform using adjacent band information. In the proposed algorithm, we divide the spectral bands into a number of groups. Then, within each group, we select a representative spectral band which has highest correlations with the remaining spectral bands within the group. The representative image is compressed using the 2-dimensional SPIHT algorithm. A characteristic of the SPIHT algorithm is that it provides information on significant coefficients of transformed images, which have large energies. We explore this characteristic and use the information to encode adjacent band images of the representative image.