In the method of combining Object detection Network (ObjectNet) and scene recognition network, the object features extracted by the ObjectNet and the scene features extracted by the scene network are inconsistent in dimensionality and property, and there is redundant information in the object features that affects the scene judgment, resulting in low recognition accuracy of scenes. To solve this problem, an improved indoor scene recognition method combined with object detection was proposed. First, the Class Conversion Matrix (CCM) was introduced into the ObjectNet to convert the object features output by ObjectNet, so that the dimension of the object features was consistent with that of the scene features, as a result, the information loss caused by inconsistency of the feature dimensions was reduced. Then, the Context Gating (CG) mechanism was used to suppress the redundant information in the features, reducing the weight of irrelevant information, and increasing the contribution of object features in scene recognition. The recognition accuracy of the proposed method on MIT Indoor67 dataset reaches 90.28%, which is 0.77 percentage points higher than that of Spatial-layout-maintained Object Semantics Features (SOSF) method; and the recognition accuracy of the proposed method on SUN397 dataset is 81.15%, which is 1.49 percentage points higher than that of Hierarchy of Alternating Specialists (HoAS) method. Experimental results show that the proposed method improves the accuracy of indoor scene recognition.

Aiming at the research of the cooperative obstacle-avoidance tracing based on traditional flocking control model，which is proposed by Reynolds and implemented by Tanner, it has been improved by us and added the Steer to Avoid obstacle avoidance method. This model has a high efficiency in avoiding convex obstacle in tracking target. If the method is applied to the environment of concave obstacles, nodes will stuck in the concave zone and could not get out, because the target has an attraction power to nodes when it comes to a Steer to Avoid judgment. This paper proposed a new model for concave obstacles by further improving the Steer to Avoid method. Temporarily cancel the attraction from the target when it comes to a concave environment judgment, and then constantly searching the path along the edge of obstacles. Finally, nodes could get out of the concave obstacles and reach target. Simulation results showed that the proposed model，while compared to the traditional model, has a marked increase on average rate and time efficiency in avoiding obstacle. Also, it can succeed in avoiding mobile concave obstacles in unknown environment.