As a key component of intelligent vehicle technology, automatic parking technology has become a popular research topic. Automatic parking technology can complete parking operations safely and quickly without a driver and can improve driving comfort, while greatly reducing the probability of parking accidents. An automatic parking system based on parking scene recognition is proposed in this paper to resolve the following issues with existing automatic parking systems: parking scene recognition methods are less intelligent, vehicle control has a low degree of automation, and the research scope is limited to traditional fuel vehicles. To increase the utilization of parking spaces and parking convenience, machine vision and pattern recognition techniques are introduced to intelligently recognize a vertical parking scenario, plan a reasonable parking path, develop a path tracking control strategy to improve the vehicle control automation, and explore a highly intelligent automatic parking technology roadmap. This paper gives three key aspects of system solutions for an automatic parking system based on parking scene recognition: parking scene recognition, parking path planning, and path tracking and control.

Obtaining information on parking slots is a prerequisite for the development of automatic parking systems, which is an essential part of the automatic driving processes. In this paper, we proposed a parking-slot-marking detection approach based on deep learning. The detection process involves the generation of mask of the marking-points by using the Mask R-CNN algorithm, extracting parking guidelines and parallel lines on the mask using the line segment detection (LSD) to determine the candidate parking slots. The experimental results show that the proposed method works well under the condition of complex illumination and around-view images from different sources, with a precision of 94.5% and a recall of 92.7%. The results also indicate that it can be applied to diverse slot types, including vertical, parallel and slanted slots, which is superior to previous methods.

The rapidly increasing connectedness of modern vehicles leads to new security challenges for Intelligent Connected Vehicle (ICV), some potential attackers perform unauthorized access to the vehicle by injecting malicious information into in-vehicle Electronic Control Units (ECUs). So, in this paper, a secure communication architecture with an attribute-isolated for ICV was proposed. Firstly, the analysis of the functional attributes of all in-vehicle ECUs in the intelligent connected environment and the division of the functional attributes of ECUs into five classifications were performed. Secondly, based on the aboveclassified attributes, a secure communication architecture with an attribute-isolated was demonstrated. ECUs have different access rights, allowing only the ECUs with the same functional attributes to communicate. After that, it was proved that the proposed architecture could resist forgery and eavesdropping attacks. Finally, the proposed architecture was constructed on the hardware environment and evaluated on an In-Vehicle Network Simulator (IVNS). The evaluation results show that the average memory usage with 120 ECUs and 100 messages is below 40 MB and the bus load can be reduced to 18.96% using the proposed architecture compared with existing architectures. Therefore, the proposed architecture solves the tradeoff between unauthorized access and the high bus load of existing architectures.

As a key technology for intelligent vehicles, automatic parking is becoming increasingly popular in the area of research. Automatic parking technology is available for safe and quick parking operations without a driver, and improving the driving comfort while greatly reducing the probability of parking accidents. An automatic parking path planning and tracking control method is proposed in this paper to resolve the following issues presented in the existing automatic parking systems, that is, low degree of automation in vehicle control; lack of conformity between segmented path planning and real vehicle motion models; and low success rates of parking due to poor path tracking. To this end, this paper innovatively proposes preview correction which can be applied to parking path planning, and detects the curvature outliers in the parking path through the preview algorithm. In addition, it is also available for correction in advance to optimize the reasonable parking path. Meanwhile, the dual sliding mode variable structure control algorithm is used to formulate path tracking control strategies to improve the path tracking control effect and the vehicle control automation. Based on the above algorithm, an automatic parking system was developed and the real vehicle test was completed, thus exploring a highly intelligent automatic parking technology roadmap. This paper provides two key aspects of system solutions for an automatic parking system, i.e., parking path planning and path tracking control.

At present, the realization of autonomous valet parking (AVP) technology does not achieve information interaction between the parking spaces and vehicles, and accurate parking spaces information perception cannot be obtained when the accuracy of the search is not precise. In addition, when using the camera vision to identify the parking spaces, traditional parking space features such as parking lines and parking angles recognition are susceptible to light and environment. Especially when the vehicle nearby partially occupies the parking space to be parked, it is not easy to determine whether it is a valid empty parking space. This paper proposes a parking space recognition method based on parking space features in the scene of AVP. By constructing the multi-dimensional features containing the parking space information, the cameras are used to extract features’ contour, locate features’ position and recognize features. In this paper, a new similarity calculation formula is proposed to recognize the stained features through template matching algorithm. According to the relative position relationship between the feature and parking space, the identification of effective empty parking spaces and their boundaries is realized. The experimental results show that compared with the recognition of traditional parking lines and parking angles, this method can identify effective empty parking spaces even when the light conditions are complex and the parking spaces are partially occupied by adjacent vehicles, which simplifies the recognition algorithm and improves the reliability of the parking spaces identification.

To make intelligent vehicles obtain human drivers' steering characteristic, a new single point preview-based human-like driver model is proposed, which contains a preview decision module and a steering wheel angle calculation module. Enlightened by the visual gaze mechanism of human drivers when passing a curved road, the preview decision module is established using Takagi-Sugeno fuzzy inference system (T-S FIS) to adaptively adjust the preview point position in both longitudinal and lateral direction, and the steering angle calculation module would use the adjusted preview point to generate steering command based on pure pursuit algorithm. Ant colony optimization (ACO) method is used to optimize the fuzzy rules in the preview decision module according to the similarity of trajectories between the proposed driver model and human drivers. The proposed human-like driver model is verified on a two-lane urban curved road. Five experienced human drivers' driving trajectories under different speeds are collected for the verification. After the preview decision module optimization done in the PreScan/Simulink simulation platform, the proposed human-like driver model shows higher similarity with experienced human drivers than the driver model with fixed preview distance or the driver model only with changeable longitudinal preview distance.