• Convergence Security Journal
• /
• v.23 no.5
• /
• pp.125-133
• /
• 2023

This paper proposes an AI model that determines parking and non-parking zones based on return authentication photos to address parking issues that may arise in shared electric scooter systems. In this study, we used a pre-trained Segformer_b0 model on ADE20K and fine-tuned it on tactile blocks and electric scooters to extract segmentation maps of objects related to parking and non-parking areas. We also presented a method to perform binary classification of parking and non-parking zones using the Swin model. Finally, after labeling a total of 1,689 images and fine-tuning the SegFomer model, it achieved an mAP of 81.26%, recognizing electric scooters and tactile blocks. The classification model, trained on a total of 2,817 images, achieved an accuracy of 92.11% and an F1-Score of 91.50% for classifying parking and non-parking areas.

• The Transactions of the Korea Information Processing Society
• /
• v.13 no.7
• /
• pp.311-318
• /
• 2024

As the amount of data increases due to the development of artificial intelligence and the Internet, data management is becoming increasingly important, and the efficient utilization of data retrieval and memory space is crucial. In this study, we investigate how to optimize search speed and memory efficiency by analyzing data structure based on metadata. As a research method, we compared and analyzed the performance of the array, association list, dictionary binary tree, and graph data structures using metadata of photographic images, focusing on temporal and space complexity. Through experimentation, it was confirmed that dictionary data structure performs best in collection speed and graph data structure performs best in search speed when dealing with large-scale image data. We expect the results of this paper to provide practical guidelines for selecting data structures to optimize search speed and memory efficiency for the images data.

• The Journal of the Korea Contents Association
• /
• v.10 no.7
• /
• pp.107-114
• /
• 2010

Color is a sense signal for human to perceive being through light, and the color is divided into chromatic color and achromatic color. Chromatic color has hue, intensity, and saturation, but achromatic color has only intensity among the properties of chromatic color and doesn't have hue and saturation. Therefore it is important to split colors of image into area for human to perceive colors and not to perceive ones based on vision of human being. In this paper, we find a function to split colors of image into chromatic region of chromatic color region and achromatic region of achromatic color region. First, the input image of RGB color space is converted into the image of HSI color space in consideration of human vision and get a binary image from the converted image. After then, a function to split colors into ROC(ROC: Region of chromatic.) and ROA(ROA:Region of achromatic) is yield. It is difficult to split color of a general image into ROC and ROA. Therefore, to get the chromatic area and achromatic area, we make gradient images to have all range of intensity and range of saturation and to have a little range of hue and yield the function. The evaluation is tested using subjective-quality by 50 non-experts for result images of test images and general images. The results of the proposed method get better 27.5~32.96% than these of the conventional method

• Journal of the Korea Society of Computer and Information
• /
• v.21 no.5
• /
• pp.31-40
• /
• 2016

In this paper, we analyse the characteristics of the edge image and propose a new entropy coding optimized to the compression of the edge image. The pixel values of the edge image have the Gaussian distribution around '0', and most of the pixel values are '0'. By using this analysis, the Zero Block technique is utilized in spatial domain. And the Intra Prediction Mode of the edge image is similar to the mode of the surrounding blocks or likely to be the Planar Mode or the Horizontal Mode. In this paper, we make use of the MPM technique that produces the Intra Prediction Mode with high probability modes. By utilizing the above properties, we design a new entropy coding method that is suitable for edge image and perform the compression. In case the existing compression techniques are applied to edge image, compression ratio is low and the algorithm is complicated as more than necessity and the running time is very long, because those techniques are based on the natural images. However, the compression ratio and the running time of the proposed technique is high and very short, respectively, because the proposed algorithm is optimized to the compression of the edge image. Experimental results indicate that the proposed algorithm provides better visual and PSNR performance up to 11 times than the JPEG.

• Ocean and Polar Research
• /
• v.35 no.3
• /
• pp.239-247
• /
• 2013

Digital image processing is a process to analyze a large volume of information on digital images. In this study, Artemia hatching rate was measured by automatically classifying and counting cysts and larvae based on color imaging data from cyst hatching experiments using an image processing technique. The Artemia hatching rate estimation consists of a series of processes; a step to convert the scanned image data to a binary image data, a process to detect objects and to extract their shape information in the converted image data, an analysis step to choose an optimal discriminant function, and a step to recognize and classify the objects using the function. The function to classify Artemia cysts and larvae is optimally estimated based on the classification performance using the areas and the plan-form factors of the detected objects. The hatching rate using the image data obtained under the different experimental conditions was estimated in the range of 34-48%. It was shown that the maximum difference is about 19.7% and the average root-mean squared difference is about 10.9% as the difference between the results using an automatic counting (this study) and a manual counting were compared. This technique can be applied to biological specimen analysis using similar imaging information.

• Journal of the Korean Institute of Telematics and Electronics D
• /
• v.35D no.9
• /
• pp.62-69
• /
• 1998

We demonstrated the character display unit using a binary phase hologram array and a liquid crystal-spatial light modulator (LC-SLM). It combines the dynamic property of the LC-SLM with the high-efficiency property of the phase hologram fabricated by photolithography. Experimental results of the proposed unit are presented. The character display unit proposed in this paper has a fundamental and important meaning as new method displaying images by using light, and it will be used in optical information processing and optical communications fields.

• Journal of Digital Contents Society
• /
• v.17 no.6
• /
• pp.461-469
• /
• 2016

A novel steganography based on ternary BCH code with t = 2 is presented in this paper. Proposed method utilizes powerful BCH code with t = 2 for data hiding to the DCT coefficients from JPEG images. The presented data hiding technique uses a proposed look up table approach for searching multiple solutions for ternary BCH code with t = 2. The proposed look up table approach enables fast and efficient search for locations of DCT coefficients, which are necessary to modify for hiding data. Presented data hiding technique is the first steganography technique based on ternary BCH code. Experimental results clearly indicate advantages of using ternary BCH compared to binary BCH.

• Journal of the Institute of Electronics Engineers of Korea SP
• /
• v.38 no.5
• /
• pp.474-485
• /
• 2001

This paper proposes a user-key-based fragile watermarking for detecting image modification. The embedding data in a form of binary image for authentication are inserted to the DCT coefficients of each block of the given image. To minimize possible exposure of being watermarked and the location of insertion, it is proposed to utilize a user-specific key in randomizing those information. Each DCT block hides one bit of data, all of which represent the user-specific authentication data. Experiments with 5 real images demonstrate that the proposed method not only detects whether there is modification or not, but also the actual location of modification with minimal visual deterioration. However, the proposed method has room for improvement against its loss of watermark by an attack of compression by more than 50%.

• Proceedings of the Korean Society of Precision Engineering Conference
• /
• 2005.06a
• /
• pp.1858-1861
• /
• 2005

Automatic nut sorting machine used to calculate the size of inserted nut and detect the black spot defection is introduced in this paper. Because most of farm products are imported from the underdeveloped countries, domestic farm products have no place to be sold in market. To overcome this critical situation, lowering the productivity cost is strongly demanded to compete with foreign corps. Imaged processed nut sorting algorithm is developed to the automatic nut sorting machine to remove the sorting time which takes lots of man power. This system is composed of mainly two parts, mechanical parts and vision system. The purpose of mechanical part is supplying the nuts automatically to make computer system capture the images of objects. Simplified mechanical system was assembled followed by 3D simulation by Pro/E design for the adaptive cost effects. Several image processing algorithms are designed to detect the spot defects and calculate the size of nuts. Test algorithm shows good results to the designed automatic nut sorting system.

• The KIPS Transactions:PartB
• /
• v.9B no.3
• /
• pp.337-342
• /
• 2002

This paper proposes a new adaptive image labeling algorithm fur object analysis of the binary images. The proposed labeling algorithm need not merge/order of complex equivalent labels like classical labeling algorithm and the processing is done during only 1 Pass. In addition, this algorithm can be extended for gray-level image easily. Experiment result with HIPR image library shows that the proposed algorithm process more than 2 times laster than compared algorithm.