Настройки

Укажите год
-

Небесная энциклопедия

Космические корабли и станции, автоматические КА и методы их проектирования, бортовые комплексы управления, системы и средства жизнеобеспечения, особенности технологии производства ракетно-космических систем

Подробнее
-

Мониторинг СМИ

Мониторинг СМИ и социальных сетей. Сканирование интернета, новостных сайтов, специализированных контентных площадок на базе мессенджеров. Гибкие настройки фильтров и первоначальных источников.

Подробнее

Форма поиска

Поддерживает ввод нескольких поисковых фраз (по одной на строку). При поиске обеспечивает поддержку морфологии русского и английского языка
Ведите корректный номера.
Ведите корректный номера.
Ведите корректный номера.
Ведите корректный номера.
Укажите год
Укажите год
Применить Всего найдено 5. Отображено 5.
16-05-2023 дата публикации

White rot strain with high lignin degradation activity and application thereof

Номер: CN116121075A
Автор: MA XINGXIA, XIONG YIXIN, FANG XUAN, LI JIAXIN
Принадлежит:

The invention discloses a white rot strain with high lignin degradation activity and application of the white rot strain in paper pulp production and wastewater decoloration. The white rot strain is trametes hirsuta XY1 with the preservation number of CGMCC (China General Microbiological Culture Collection Center) No. 40195. The strain can selectively degrade lignin, has the high-activity secretion characteristic of laccase (Lac), manganese peroxidase (Mnp) and lignin peroxidase (Lip), and has the characteristics of low amylase activity and low xylanase activity. After eucalyptus is directly treated by the trametes hirsutum strain fermentation liquor without chemical treatment modes such as acid or alkali, eucalyptus chips can obtain high pulp yield, the quality of prepared coarse pulp is improved, and paper prepared from the paper pulp is improved in flexibility and whiteness. When the strain fermentation liquor is used for treating papermaking wastewater, the decolorization rate reaches ...

Подробнее
Номер записи: 1
25-04-2023 дата публикации

Adaptive false positive pulmonary nodule elimination method based on deep learning

Номер: CN116012355A
Автор: WANG YUJUAN, DENG LEI, XIONG YIXIN, SONG YONGDUAN, XIANG JINGXUAN, WANG YUFAN, JIN QIJIAN, CHEN QING, WANG CHAO, DAI YUNFEI
Принадлежит:

The invention discloses a deep learning-based adaptive false positive pulmonary nodule elimination method, which is implemented by a 3DCNN-based adaptive false positive elimination network, can classify pulmonary nodule candidate blocks with different sizes, and eliminates false positive pulmonary nodules. In addition, the adaptive rate is automatically adjusted according to the diameter of the pulmonary nodule. According to the method, to-be-detected nodules are sampled to 22 * 22 * 20 and 32 * 32 * 24 sizes and then are respectively put into two independent three-dimensional residual convolutional neural networks for training and classification, outputs of the two independent three-dimensional residual convolutional neural networks are fused by means of an adaptive rate, nodule candidate blocks of different sizes can be classified, and false positive nodules are eliminated; according to the method, the diameter of the nodule is fully considered, the problem that other systems are poor ...

Подробнее
Номер записи: 2
06-06-2023 дата публикации

Pulmonary nodule detection and elimination method based on deep learning

Номер: CN116228685A
Автор: WANG YUJUAN, XIONG YIXIN, DENG LEI, SONG YONGDUAN, JIN QIJIAN, PAN YANRUI, CHEN QING, DAI YUNFEI, WANG CHAO
Принадлежит:

The invention discloses a pulmonary nodule detection and elimination method based on deep learning, and the method comprises the steps: carrying out the fusion before a network generates a result, keeping the same weight for the prediction result of each model, and obtaining a final result through a weighted frame fusion postprocessing method. The method for judging the continuity of the nodules according to the intersection-to-union ratio of the continuous pictures is provided, the nodules can be quickly converted from two dimensions to three dimensions, and misjudgment can be effectively reduced. According to the method, the pulmonary nodules are detected based on the two-dimensional target detection model, more calculation power is saved, the three-dimensional reconstruction technology is combined after detection is completed, quicker and more accurate pulmonary nodule detection is achieved, in addition, the Sa-Yolo model is provided, and a segmentation attention mechanism is introduced ...

Подробнее
Номер записи: 3
30-05-2023 дата публикации

Pulmonary nodule image detection method based on deep learning

Номер: CN116188404A
Автор: SONG YONGDUAN, DENG LEI, XIONG YIXIN, ZHOU YONGCHENG, SHEN ZHIXI, JIN QIJIAN
Принадлежит:

The invention discloses a pulmonary nodule detection method based on deep learning. The pulmonary nodule detection method comprises the following steps of target detection of pulmonary nodules, edge segmentation of the pulmonary nodules and three-dimensional reconstruction of the pulmonary nodules. According to the pulmonary nodule detection method based on deep learning, a rapid two-stage nodule detection, segmentation and classification CAD system based on 2DCNN is provided, the malignant degree of the nodule can be detected, and the morphological characteristics of the nodule can also be detected; a segmentation attention module is introduced, the interpretability of the model is improved, and feature extraction of small nodules is enhanced; an original backbone network is replaced by a lightweight network, so that the network is simplified, the model parameter quantity is reduced, and the calculation cost is reduced; after the complete nodule slices are obtained, the nodules to be trained ...

Подробнее
Номер записи: 4
01-08-2023 дата публикации

Switch detection system and method under complex condition based on deep learning

Номер: CN116524270A
Автор: XIONG YIXIN, HE XIANG, LI CHEN, WANG MINJIE, SUN XILIANG, GUO YANMING
Принадлежит:

The invention discloses a deep learning-based switch detection system and method under a complex condition, and the method comprises the steps: obtaining an initial image and a historical image through a camera, and carrying out the labeling of the historical image, and obtaining a labeled image; training the anchor-frame-free real-time target detection model by using a historical image to obtain a trained anchor-frame-free real-time target detection model; inputting the initial image into a trained anchor-frame-free real-time target detection model, and outputting to obtain a switch position; intercepting a switch image corresponding to the switch position from the initial image; training the on-off state detection model by using the annotated image to obtain a trained on-off state detection model; inputting the switch image into a switch state detection model, and outputting to obtain a switch state corresponding to the switch position; and finally, the switch state and the switch position ...

Подробнее
Номер записи: 5