bIn this study, a 96-h laboratory reduction test was conducted with strain BDHSH06 (GenBank accession no. EF011103) as the test strain for Bdellovibrio and like organisms (BALOs) and 20 susceptible marine bacterial strains forming microcosms as the targets. The results showed that BDHSH06 reduced the levels of approximately 50% of prey bacterial strains within 96 h in the seawater microcosms. An 85-day black tiger shrimp (Penaeus monodon) rearing experiment was performed. The shrimp survival rate, body length, and weight in the test tanks were 48.1% ؎ 1.2%, 99.8 ؎ 10.0 mm, and 6.36 ؎ 1.50 g, respectively, which were values significantly (P < 0.05) higher than those for the control, viz., 31.0% ؎ 2.1%, 86.0 ؎ 11.1 mm, and 4.21 ؎ 1.56 g, respectively. With the addition of BDHSH06, total bacterial and Vibrio numbers were significantly reduced (P < 0.05) by 1.3 to 4.5 log CFU · ml ؊1 and CFU · g ؊1 in both water and shrimp intestines, respectively, compared to those in the control. The effect of BDHSH06 on bacterial community structures in the rearing water was also examined using PCR amplification of the 16S rRNA gene and denaturing gradient gel electrophoresis (DGGE). The DGGE profiles of rearing water samples from the control and test tanks revealed that the amounts of 44% of the bacterial species were reduced when BDHSH06 was added to the rearing water over the 85-day rearing period, and among these, approximately 57.1% were nonculturable. The results of this study demonstrated that BDHSH06 can be used as a biocontrol/probiotic agent in P. monodon culture.
Scene text detection is to detect the position of a text in the natural scene, the quality of which will directly affect the subsequent text recognition. It plays an important role in fields such as image retrieval and autopilot. How to perform multi-scale and multi-oriented text detection in the scene still remains as a problem. This paper proposes an effective scene text detection method that combines the convolutional neural network (CNN) and recurrent neural network (RNN). In order to better adapt to texts in different scales, feature pyramid networks (FPN) have been applied in the CNN part to extract multi-scale features of the image. We then utilize bidirectional long-short-term memory (Bi-LSTM) to encode these features to make full use of the text sequence characteristics with the outputs as a series of text proposals. The generated proposals are finally linked into a text line through a well-designed text connector, which can be flexibly adapted to any oriented texts. The proposed method is evaluated on three public datasets: ICDAR2013, ICDAR2015, and USTB-SV1K. For ICDAR2013 and USTB-1K, we have reached 92.5% and 62.6% F-measure, respectively. Our method has reached 72.8% F-measure on the more challenging ICDAR2015 which demonstrates the effectiveness of our method. INDEX TERMS Scene text detection, multi-orientation, convolutional neural network, recurrent neural network, residual network.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.