• The Journal of Korean Institute of Information Technology
• /
• v.16 no.11
• /
• pp.85-95
• /
• 2018

With the popularity of smartphones, most peoples have been using mobile cameras to capture photographs. However, due to insufficient amount of lights in a low lighting condition, unwanted noises can be generated during image acquisition. To remove the noise, a method of using deep convolutional neural networks is introduced. However, this method still lacks the ability to describe textures and edges, even though it has made significant progress in terms of visual quality performance. Therefore, in this paper, the HOG (Histogram of Oriented Gradients) images that contain information about edge orientations are used. More specifically, a method of learning deep convolutional neural networks is proposed by stacking noise and HOG images into an input tensor. Experiment results confirm that the proposed method not only can obtain excellent result in visual quality evaluations, compared to conventional methods, but also enable textures and edges to be improved visually.

• Archives of Plastic Surgery
• /
• v.47 no.2
• /
• pp.171-177
• /
• 2020

Background Infrabrow blepharoplasty has become a common surgical method used to rejuvenate aged upper eyelids in Asians. In this paper, we describe the parallel excision method for infrabrow blepharoplasty as a useful alternative to the conventional elliptical excision method. The authors' experience over a 3-year period is presented and reviewed. Methods A retrospective review of parallel excision infrabrow blepharoplasty cases at our hospital between 2014 and 2017 was performed. Three oculoplastic surgeons compared preoperative and postoperative photographs using the Strasser grading system. Results From the medical records of 123 patients, a total of 93 patients with moderate-to-severe bilateral dermatochalasis were selected as subjects. The exclusion criterion was levator function less than 8 mm. The total mean follow-up period was 2 years (range, 0.5-3.5 years). The mean skin excision height and width were 9.75 mm (range, 5-16 mm) and 58.51 mm (range, 42-75 mm), respectively. All patients who underwent surgery recovered without major complications, and all patients had high levels of satisfaction and improvements in their visual field. In the Strasser evaluation performed by the oculoplastic surgeons, most patients were found to have excellent results. Conclusions The parallel excision method for infrabrow blepharoplasty is a safe and effective technique that yields more natural- and youthful-looking eyelids than the conventional elliptical excision method. In our method, more effective manipulation of the orbicularis oculi muscle led to a reduction in frontalis compensation, resolution of sunken eyelids, and correction of lateral hooding.

• Journal of the Society of Cosmetic Scientists of Korea
• /
• v.41 no.2
• /
• pp.97-103
• /
• 2015

Until recently, the three conventional evaluation methods, which are instrumental (Chromameter$^{(R)}$ CR-400 and Mexameter$^{(R)}$ M18) and visual assessments have been used frequently for skin color evaluation. However, we took notice the potential of image analysis as a new tool to evaluate color change of skin. To reveal the reliability of the image analysis for the evaluation of whitening agents, 34 healthy female volunteers with hyperpigmentation were recruited, and the selected volunteers applied the whitening products containing Vitamin C twice a day in the morning and evening and received iontophoresis treatments once a week for 8 weeks. The changes in hyperpigmentation evaluated by Chromameter$^{(R)}$, Mexameter$^{(R)}$ and visual assessment were compared with the results from the image analysis. As with $L^*$ value trends of the analysis using Chromameter$^{(R)}$, the V value from the image analysis increased after applying the test products compared with baseline values. Furthermore, V value showed a positive correlation with $L^*$ value (r = 0.494, p < 0.01) and negative correlation with MI (r = - 0.683, p < 0.01) and VG (r = - 0.549, p < 0.01). Therefore, image analysis may be considered as an effective method to complement the limitations of visual assessment for whitening efficacy in Asians.

• Proceedings of the KIEE Conference
• /
• 2004.11c
• /
• pp.107-109
• /
• 2004

In this paper, authors propose the sensor fusion system that can recognize multiple 3D objects from 2D projection images and tactile information. The proposed system focuses on improving recognition performance of 3D object. Unlike the conventional object recognition system that uses image sensor alone, the proposed method uses tactual sensors in addition to visual sensor. Neural network is used to fuse these informations. Tactual signals are obtained from the reaction force by the pressure sensors at the fingertips when unknown objects are grasped by four-fingered robot hand. The experiment evaluates the recognition rate and the number of teaming iterations of various objects. The merits of the proposed systems are not only the high performance of the learning ability but also the reliability of the system with tactual information for recognizing various objects even though visual information has a defect. The experimental results show that the proposed system can improve recognition rate and reduce learning time. These results verify the effectiveness of the proposed sensor fusion system as recognition scheme of 3D object.

• Journal of Korea Multimedia Society
• /
• v.19 no.5
• /
• pp.837-847
• /
• 2016

Tone mapping for High Dynamic Range(HDR) image provides matching human visual perception between real world scene and displayable devices. Recently, a tone mapping algorithm based on localized gamma correction is proposed. This algorithm is using human visual properties of contrast and colorfulness with background intensity, generating a weight map for gamma correction. However, this method have limitations of controlling enhancement region as well as generating halo artifacts caused by the weight map construction. To overcome aforementioned limitations, proposed algorithm in this paper modifies previous weight map, considering base layer intensity of input luminance channel. By determining enhancement region locally and globally based on base layer intensity, gamma values are corrected accordingly. Therefore, proposed algorithm selectively enhances local brightness and controls strength of edges. Subjective evaluation using z-score shows that our proposed algorithm outperforms the conventional methods.

• Journal of IKEEE
• /
• v.11 no.4
• /
• pp.219-226
• /
• 2007

An efficient block based two-layer JPEG encoding algorithm is proposed to compress high dynamic range (HDR) images in this work. The proposed algorithm separates an input HDR image into a tone-mapped low dynamic range (LDR) image and a ratio image, which represents the quotients of the original HDR pixels divided by the tone-mapped LDR pixels. Then, the tone-mapped LDR image is compressed using the standard JPEG scheme to preserve backward compatibility and the ratio image is encoded to minimize a cost function that models the perception of each block with different quantization parameters in the human visual system (HVS). Simulation results show that the proposed algorithm provides better performance than the conventional method, which encodes the ratio image without any prior information of blocks.

• Journal of Biomedical Engineering Research
• /
• v.26 no.3
• /
• pp.129-132
• /
• 2005

It is one of the most important issues to determine a target brain image that gives a common coordinate system for a constructing population-based brain atlas. The purpose of this study is to provide a simple and reliable procedure that determines the target brain image among the group based on the inherent structural information of three-dimensional magnetic resonance (MR) images. It uses only 11 lines defined automatically as a feature vector representing structural variations based on the Talairach coordinate system. Average characteristic vector of the group and the difference vectors of each one from the average vector were obtained. Finally, the individual data that had the minimum difference vector was determined as the target. We determined the target brain image by both our algorithm and conventional visual inspection for 20 healthy young volunteers. Eighteen fiducial points were marked independently for each data to evaluate the similarity. Target brain image obtained by our algorithm showed the best result, and the visual inspection determined the second one. We concluded that our method could be used to determine an appropriate target brain image in constructing brain atlases such as disease-specific ones.

• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.54 no.3
• /
• pp.156-165
• /
• 2005

Human being recognizes the physical world by integrating a great variety of sensory inputs, the information acquired by their own action, and their knowledge of the world using hierarchically parallel-distributed mechanism. In this paper, authors propose the sensor fusion system that can recognize multiple 3D objects from 2D projection images and tactile informations. The proposed system focuses on improving recognition performance of 3D objects. Unlike the conventional object recognition system that uses image sensor alone, the proposed method uses tactual sensors in addition to visual sensor. Neural network is used to fuse the two sensory signals. Tactual signals are obtained from the reaction force of the pressure sensors at the fingertips when unknown objects are grasped by four-fingered robot hand. The experiment evaluates the recognition rate and the number of learning iterations of various objects. The merits of the proposed systems are not only the high performance of the learning ability but also the reliability of the system with tactual information for recognizing various objects even though the visual sensory signals get defects. The experimental results show that the proposed system can improve recognition rate and reduce teeming time. These results verify the effectiveness of the proposed sensor fusion system as recognition scheme for 3D objects.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.15 no.2
• /
• pp.421-441
• /
• 2021

Fruit detection in orchards is one of the most crucial tasks for designing the visual system of an automated harvesting robot. It is the first and foremost tool employed for tasks such as sorting, grading, harvesting, disease control, and yield estimation, etc. Efficient visual systems are crucial for designing an automated robot. However, conventional fruit detection methods always a trade-off with accuracy, real-time response, and extensibility. Therefore, an improved method is proposed based on coarse-to-fine multitask cascaded convolutional networks (MTCNN) with three aspects to enable the practical application. First, the architecture of Fruit-MTCNN was improved to increase its power to discriminate between objects and their backgrounds. Then, with a few manual labels and operations, synthetic images and labels were generated to increase the diversity and the number of image samples. Further, through the online hard example mining (OHEM) strategy during training, the detector retrained hard examples. Finally, the improved detector was tested for its performance that proved superior in predicted accuracy and retaining good performances on portability with the low time cost. Based on performance, it was concluded that the detector could be applied practically in the actual orchard environment.

• Journal of Mechanical Science and Technology
• /
• v.18 no.6
• /
• pp.961-971
• /
• 2004

Methods for determination of the crack opening stress intensity factor ($K_{op}$) and for estimation of the effective stress intensity factor range ($\Delta{K}_{eff}$) are evaluated for crack growth test data of aluminum alloys. Three methods of determining $K_{op}$, visual measurement, ASTM offset compliance method, and the neural network method proposed by Kang and Song, and three methods of estimating $\Delta{K}_{eff}$, conventional, the 2/PIO and 2/PI methods proposed by Donald and Paris, are compared in a quantitative manner by using evaluation criteria. For all $K_{op}$ determination methods discussed, the 2/PI method of estimating $\Delta{K}_{eff}$ provides good results. The neural network method of determining $K_{op}$ provides good correlation of crack growth data. It is recommended to use 2/PI estimation with the neural $K_{op}$ determination method. The ASTM offset method used in conjunction with 2/PI estimation shows a possibility of successful application. It is desired to improve the ASTM method.