• 대한원격탐사학회지
• /
• 제40권1호
• /
• pp.93-101
• /
• 2024

Accurate field crop classification is essential for various agricultural applications, yet existing methods face challenges due to diverse crop types and complex field conditions. This study aimed to address these issues by combining support vector machine (SVM) models with multi-seasonal unmanned aerial vehicle (UAV) images, texture information extracted from Gray Level Co-occurrence Matrix (GLCM), and RGB spectral data. Twelve high-resolution UAV image captures spanned March-October 2021, while field surveys on three dates provided ground truth data. We focused on data from August (-A), September (-S), and October (-O) images and trained four support vector classifier (SVC) models (SVC-A, SVC-S, SVC-O, SVC-AS) using visual bands and eight GLCM features. Farm maps provided by the Ministry of Agriculture, Food and Rural Affairs proved efficient for open-field crop identification and served as a reference for accuracy comparison. Our analysis showcased the significant impact of hyperparameter tuning (C and gamma) on SVM model performance, requiring careful optimization for each scenario. Importantly, we identified models exhibiting distinct high-accuracy zones, with SVC-O trained on October data achieving the highest overall and individual crop classification accuracy. This success likely stems from its ability to capture distinct texture information from mature crops.Incorporating GLCM features proved highly effective for all models,significantly boosting classification accuracy.Among these features, homogeneity, entropy, and correlation consistently demonstrated the most impactful contribution. However, balancing accuracy with computational efficiency and feature selection remains crucial for practical application. Performance analysis revealed that SVC-O achieved exceptional results in overall and individual crop classification, while soybeans and rice were consistently classified well by all models. Challenges were encountered with cabbage due to its early growth stage and low field cover density. The study demonstrates the potential of utilizing farm maps and GLCM features in conjunction with SVM models for accurate field crop classification. Careful parameter tuning and model selection based on specific scenarios are key for optimizing performance in real-world applications.

• 한국멀티미디어학회논문지
• /
• 제17권3호
• /
• pp.277-286
• /
• 2014

본 논문에서는 Gabor 특징과 MDLC 특징, 그리고 co-occurrence 특징의 융합에 의한 질감 특징 기반언어 인식 방법을 제안한다. 제안된 방법에서는 먼저 시험 영상에 Gabor 변환에 이은 크기 연산자를 적용하여 Gabor 크기 영상을 얻고 그 통계치를 계산하여 결과를 벡터화한다. 이어서 MDLC 연산자를 이용하여 MDLC 영상을 얻고 역시 그 통계치를 계산하여 벡터화한다. 다음으로 시험 영상으로부터 GLCM을 계산하고 이를 이용하여 co-occurrence 특징을 계산한 다음 벡터화한다. 이들 Gabor, MDLC, co-occurrence 특징에 의한 벡터들은 벡터 융합에 의하여 특징 벡터로 사용된다. 분류 단계에서는 얼굴 인식에 주로 사용되는 WPCA를 분류기로 하여 시험 특징 벡터와 가장 유사한 학습 특징 벡터를 찾는다. 제안된 방법의 성능은 15개국 언어의 문서를 스캔하여 얻은 시험 문서 영상 DB에 대한 평균 인식률을 조사하여 알아본다. 실험 결과 제안된 방법은 시험 DB에 대하여 비교적 낮은 특징 벡터 차원으로 매우 우수한 언어 인식 성능을 보여준다.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제11권1호
• /
• pp.360-373
• /
• 2017

This paper presents a novel digital video steganalysis scheme against the spatial domain video steganography technology based on a spatial temporal detector (ST_D) that considers both spatial and temporal redundancies of the video sequences simultaneously. Three descriptors are constructed on XY, XT and YT planes respectively to depict the spatial and temporal relationship between the current pixel and its adjacent pixels. Considering the impact of local motion intensity and texture complexity on the histogram distribution of three descriptors, each frame is segmented into non-overlapped blocks that are $8{\times}8$ in size for motion and texture analysis. Subsequently, texture and motion factors are introduced to provide reasonable weights for histograms of the three descriptors of each block. After further weighted modulation, the statistics of the histograms of the three descriptors are concatenated into a single value to build the global description of ST_D. The experimental results demonstrate the great advantage of our features relative to those of the rich model (RM), the subtractive pixel adjacency model (SPAM) and subtractive prediction error adjacency matrix (SPEAM), especially for compressed videos, which constitute most Internet videos.

• Journal of Radiation Protection and Research
• /
• 제48권3호
• /
• pp.144-152
• /
• 2023

Background: Color texture analysis was applied as a tool for quantitative evaluation of radiation-induced skin injuries. Materials and Methods: We prospectively selected 20 breast cancer patients who underwent whole-breast radiotherapy after breast-conserving surgery. Color images of skin surfaces for irradiated breasts were obtained by using a mobile skin analyzer. The first skin measurement was performed before the first fraction of radiotherapy, and the subsequent measurement was conducted approximately 10 days after the completion of the entire series of radiotherapy sessions. For comparison, color images of the skin surface for the unirradiated breasts were measured similarly. For each color image, six co-occurrence matrices (red-green [RG], red-blue [RB], and green-blue [GB] from color channels, red [R], green [G], blue [B] from gray channels) can be generated. Four textural features (contrast, correlation, energy, and homogeneity) were calculated for each co-occurrence matrix. Finally, several statistical analyses were used to investigate the performance of the color textural parameters to objectively evaluate the radiation-induced skin damage. Results and Discussion: For the R channel from the gray channel, the differences in the values between the irradiated and unirradiated skin were larger than those of the G and B channels. In addition, for the RG and RB channels, where R was considered in the color channel, the differences were larger than those in the GB channel. When comparing the relative values between gray and color channels, the 'contrast' values for the RG and RB channels were approximately two times greater than those for the R channel for irradiated skin. In contrast, there were no noticeable differences for unirradiated skin. Conclusion: The utilization of color texture analysis has shown promising results in evaluating the severity of skin damage caused by radiation. All textural parameters of the RG and RB co-occurrence matrices could be potential indicators of the extent of skin damage caused by radiation.

• 한국멀티미디어학회논문지
• /
• 제18권8호
• /
• pp.901-908
• /
• 2015

Methods of detecting, describing, matching image features, like corners and blobs, have been actively studied as a fundamental step for image processing and computer vision applications. As one of feature description/matching methods, LLAH(Locally Likely Arrangement Hashing) describes image features based on the geometric relationship between their neighbors, and thus is suitable for scenes with poor texture. This paper presents a modified LLAH algorithm, which includes the image features themselves for robustly describing the geometric relationship unlike the original LLAH, and employes a voting-based feature matching scheme that makes feature description much simpler. Then, this paper quantitatively analyzes its performance with synthetic images in the presence of Gaussian noise.

• Korean Journal of Radiology
• /
• 제21권7호
• /
• pp.779-792
• /
• 2020

Recent advances in computer technology have generated a new area of research known as radiomics. Radiomics is defined as the high throughput extraction and analysis of quantitative features from imaging data. Radiomic features provide information on the gray-scale patterns, inter-pixel relationships, as well as shape and spectral properties of radiological images. Moreover, these features can be used to develop computational models that may serve as a tool for personalized diagnosis and treatment guidance. Although radiomics is becoming popular and widely used in oncology, many problems such as overfitting and reproducibility issues remain unresolved. In this review, we will outline the steps of radiomics used for oncology, specifically addressing applications for breast cancer patients and focusing on technical issues.

• 한국축산식품학회지
• /
• 제38권2호
• /
• pp.362-375
• /
• 2018

This study proposed a rapid microscopic examination method for pork freshness evaluation by using the self-assembled hyperspectral microscopic imaging (HMI) system with the help of feature extraction algorithm and pattern recognition methods. Pork samples were stored for different days ranging from 0 to 5 days and the freshness of samples was divided into three levels which were determined by total volatile basic nitrogen (TVB-N) content. Meanwhile, hyperspectral microscopic images of samples were acquired by HMI system and processed by the following steps for the further analysis. Firstly, characteristic hyperspectral microscopic images were extracted by using principal component analysis (PCA) and then texture features were selected based on the gray level co-occurrence matrix (GLCM). Next, features data were reduced dimensionality by fisher discriminant analysis (FDA) for further building classification model. Finally, compared with linear discriminant analysis (LDA) model and support vector machine (SVM) model, good back propagation artificial neural network (BP-ANN) model obtained the best freshness classification with a 100 % accuracy rating based on the extracted data. The results confirm that the fabricated HMI system combined with multivariate algorithms has ability to evaluate the fresh degree of pork accurately in the microscopic level, which plays an important role in animal food quality control.

• 한국컴퓨터정보학회논문지
• /
• 제16권2호
• /
• pp.249-256
• /
• 2011

텍스처 영상을 표현할 수 있는 새로운 특징(feature)의 개발은 텍스처 영상 연구에서 가장 중요한 분야 중 하나이다. 모든 종류의 텍스처 영상을 표현하는 완벽한 특징의 집합을 정립하는 것은 불가능하여 텍스처 영상을 분석하는 분야에 따라 적절한 텍스처 특징을 사용하여야 하기 때문이다. 본 논문의 첫 번째 연구 목적은 텍스처 영상을 분할하는데 유용한 새로운 텍스처 특징을 찾아내는 것이다. 실험 텍스처를 통해 본 논문에서 제안하는 새로운 특징의 영상 분할에서의 유용성 검증도 제시하였다. 한편 본 논문에서 제안한 텍스처 특징은 페이즈의 미분 데이터이다. 페이즈 정보는 텍스처에 가버 함수를 필터로 적용한 결과 영상으로부터 구해지는데, 페이즈 정보는 지금까지 텍스처 특징으로 사용하기에는 적절하지 않다고 알려져 왔으며 관련 연구도 미미한 상태이다. 그 이유는 페이즈 정보를 계산하는 방법으로 역탄젠트 오퍼레이터를 적용하여 구하는데, 이렇게 구한 페이즈 값은 $-{\pi}/2$와 ${\pi}/2$ 범위에서 정의 되는 불연속 값들이기 때문이다. 이러한 불연속적인 페이즈 데이터는 텍스처에 대한 정보를 적절하게 표현하지 못하고 왜곡된 정보를 만든다. 물론 불연속적인 페이즈 데이터를 연속적인 페이즈 데이터로 변환하는 언랩핑은 이론적으로는 불연속의 문제를 해결하지만 실제적인 적용에서는 문제점을 노출시킨다. 즉 실수와 허수 부문이 모두 0인 지점을 중심으로 페이즈 불연속을 해결하는 과정에 따라 상이한 페이즈 데이터의 값들이 도출되어 정보가 왜곡된다. 본 논문의 두 번째 연구 목적은 이러한 단점을 근본적으로 회피하는 다른 방법을 찾는 것이다. 본 연구에서 제안하는 새로운 방법은 불연속적인 페이즈를 왜곡되고 복잡한 언래핑을 거치지 않고 바로 미분함으로써 영상분할에 필요한 데이터를 도출하는 것이다. 페이즈 미분 방식은페이즈 언랩핑 절차를 필요로 하지 않기 때문에 언랩핑의 왜곡을 회피하고, 알고리즘이 매우 간단해지며, 가버함수에 적용하는 필터도 자유롭게 선택할 수 있다는 장점을 가진다. 본 논문에서는 제시하는 방법의 유용성을 검증하기 위해 가버 필터를 적용하여 도출된 결과 영상으로부터 새롭게 제시한 방법으로 계산한 페이즈로 구성된 특징 벡터를 구해 텍스처 영상의 분할에 이용하였다. 실험 결과를통해 본 논문에서 제시한 새로운 방식이 텍스처 영상분할에 필요한 유용한 정보를 생성하는 것을 확인할 수 있었다.

• 대한전자공학회논문지SP
• /
• 제48권4호
• /
• pp.76-85
• /
• 2011

본 논문에서는 Gabor 특징과 웨이브렛 영역의 BDIP와 BVLC 특징을 이용한 질감 특징 기반 언어 인식 방법을 제안한다. 제안된 방법에서는 먼저 시험 영상에 Gabor 변환과 웨이브렛 변환을 적용한다. 웨이브렛 영역의 상세 대역에는 Donoho의 연역치화를 적용하여 잡음을 제거한다. 이어서 Gabor 영상에는 크기 연산자를 적용하고 웨이브렛 부대역에는 BDIP와 BVLC 연산자를 적용한다. 그런 다음 Gabor 크기 영상과 BDIP, BVLC 부대역에 대하여 통계치를 계산하여 그 결과들을 벡터화하고 융합하여 특징 벡터로 사용한다. 분류 단계에서는 얼굴 인식에 주로 사용되는 WPCA를 분류기로 하여 시험 특징 벡터와 가장 유사한 학습 특징 벡터를 찾는다. 실험 결과 제안된 방법은 실험 문서 영상 DB에 대하여 비교적 낮은 특징 벡터 차원으로 매우 우수한 언어 인식 성능을 보여준다.

• Journal of Information Processing Systems
• /
• 제13권5호
• /
• pp.1372-1381
• /
• 2017

In this paper, a texture feature extraction method using local energy and local correlation of Gabor transformed images is proposed and applied to an image retrieval system. The Gabor wavelet is known to be similar to the response of the human visual system. The outputs of the Gabor transformation are robust to variants of object size and illumination. Due to such advantages, it has been actively studied in various fields such as image retrieval, classification, analysis, etc. In this paper, in order to fully exploit the superior aspects of Gabor wavelet, local energy and local correlation features are extracted from Gabor transformed images and then applied to an image retrieval system. Some experiments are conducted to compare the performance of the proposed method with those of the conventional Gabor method and the popular rotation-invariant uniform local binary pattern (RULBP) method in terms of precision vs recall. The Mahalanobis distance is used to measure the similarity between a query image and a database (DB) image. Experimental results for Corel DB and VisTex DB show that the proposed method is superior to the conventional Gabor method. The proposed method also yields precision and recall 6.58% and 3.66% higher on average in Corel DB, respectively, and 4.87% and 3.37% higher on average in VisTex DB, respectively, than the popular RULBP method.