Hierarchical Clustering-Based Image Retrieval for Indoor Visual Localization

Feng, Guanyuan; Jiang, Zhengang; Tan, Xiaoyang; Cheng, Feihao

doi:10.3390/electronics11213609

Cited by 6 publications

(2 citation statements)

References 61 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…At present, the mainstream visual localization methods are divided into direct 2D-3D matching methods [16,17], image retrieval methods [18], and learning-based regression methods [19]. However, the direct 2D-3D matching method is susceptible to changes in light and angle, while the learning-based regression method requires a large amount of data and extensive computation.…”

Section: Visual Localization Based On Image Retrievalmentioning

confidence: 99%

Target Localization Method Based on Image Degradation Suppression and Multi-Similarity Fusion in Low-Illumination Environments

Tang

Qin

Yang

et al. 2023

IJGI

View full text Add to dashboard Cite

Frame buildings as important nodes of urban space. The include high-speed railway stations, airports, residences, and office buildings, which carry various activities and functions. Due to illumination irrationality and mutual occlusion between complex objects, low illumination situations frequently develop in these architectural environments. In this case, the location information of the target is difficult to determine. At the same time, the change in the indoor electromagnetic environment also affects the location information of the target. Therefore, this paper adopts the vision method to achieve target localization in low-illumination environments by feature matching of images collected in the offline state. However, the acquired images have serious quality degradation problems in low-illumination conditions, such as low brightness, low contrast, color distortion, and noise interference. These problems mean that the local features in the collected images are missing, meaning that they fail to achieve a match with the offline database images; as a result, the location information of the target cannot be determined. Therefore, a Visual Localization with Multiple-Similarity Fusions (VLMSF) is proposed based on the Nonlinear Enhancement And Local Mean Filtering (NEALMF) preprocessing enhancement. The NEALMF method solves the problem of missing local features by improving the quality of the acquired images, thus improving the robustness of the visual positioning system. The VLMSF method solves the problem of low matching accuracy in similarity retrieval methods by effectively extracting and matching feature information. Experiments show that the average localization error of the VLMSF method is only 8 cm, which is 33.33% lower than that of the Kears-based VGG-16 similarity retrieval method. Meanwhile, the localization error is reduced by 75.76% compared with the Perceptual hash (Phash) retrieval method. The results show that the method proposed in this paper greatly alleviates the influence of low illumination on visual methods, thus helping city managers accurately grasp the location information of targets under complex illumination conditions.

show abstract

Section: Visual Localization Based On Image Retrievalmentioning

confidence: 99%

Target Localization Method Based on Image Degradation Suppression and Multi-Similarity Fusion in Low-Illumination Environments

Tang

Qin

Yang

et al. 2023

IJGI

View full text Add to dashboard Cite

show abstract

“…These distinctions are important in feature extraction. Image retrieval aims to teach the computer to recognize similar images [34]. As a result, the image must be preprocessed to reduce noise, lighting, and any other unnecessary differences.…”

Section: A Collecting Images Of Historical Buildingmentioning

confidence: 99%

Optimization of Historic Buildings Recognition: CNN Model and Supported by Pre-processing Methods

Rangkuti,

Hasbi Athala,

Haridhi Indallah

et al. 2023

JOIV : Int. J. Inform. Visualization

View full text Add to dashboard Cite

Several cities in Indonesia, such as Cirebon, Bandung, and Bogor, have several historical buildings that date back to the Dutch colonial period. Several Dutch colonial heritage buildings can be found in several areas. The existence of historical buildings also would attract foreign or local tourists who visit one of an area. We need a technology or model that would support the recognition and identification of buildings, including their characteristics. However, recognizing and identifying them is a problem in itself, so technology would be needed to help them. The technology or model that would be implemented in this research is the Convolutional Neural Network model, a derivative of Artificial Intelligent technology focused on image processing and pattern recognition. This process consists of several stages. The initial stage uses the Gaussian Blur, SuCK, and CLAHE methods which are useful for image sharpening and recognition. The second process is feature extraction of the image characteristics of buildings. The results of the image process will support the third process, namely the image retrieval process of buildings based on their characteristics. Based on these three main processes, they would facilitate and support local and foreign tourists to recognize historic buildings in the area. In this experiment, the Euclidean distance and Manhattan distance methods were used in the retrieval process. The highest accuracy in the retrieval process for the feature extraction process with the DenseNet 121 model with the initial process is Gaussian Blur of 88.96% and 88.46%, with the SuCK method of 88.3 and 87.8%, and with CLAHE of 87.7%, and 87.6%. We hope that this research can be continued to identify buildings with more complex characteristics and models.

show abstract