In-vehicle image processing method and image processing apparatus

An in-vehicle image processing apparatus capable of being linked to multiple cameras mounted to a vehicle, the image processing apparatus comprising:

• an image memory that stores multiple image data captured by each of the multiple cameras;
• a camera-image acquisition unit that inputs and stores the multiple image data captured by each of the cameras into the image memory;
• a projection surface geometry manager that manages projection surface geometry information including a set of three-dimensional coordinates for a projection surface that is virtually formed to surround the vehicle;
• a viewing movement controller that determines viewing information so that a viewpoint and a viewing direction relative to the projection surface are continuously transited according to given viewing transition regulations;
• an image converter that projects and combines each of the image data stored in the image memory on the projection surface, and converts the combined image data into two-dimensional image data based on the viewing information determined by the viewing movement controller;
• a display unit that displays the two-dimensional image data converted by the image converter;
• a periodic rendering trigger generator that generates a trigger signal to make the image converter start an image conversion process;
• and
• a viewing transition-regulation memory that stores the viewing transition regulations, wherein
• the viewing transition regulations stored in the viewing transition-regulation memory include start-point viewing information, end-point viewing information and viewing interpolation regulations,
• the viewing movement controller continuously changes the viewing information based on the start-point viewing information, the end-point viewing information, and the viewing interpolation regulations in the viewing transition-regulation memory, and
• the viewing interpolation regulations stored in the viewing transition-regulation memory continuously increases a transition speed for a given period of time after the transition of viewing information is started from the start-point viewing information, and continuously decreases the transition speed for a given period of time before the transition of viewing information is stopped at the end-point viewing information.

An image processing method for in-vehicle image processing apparatus capable of being linked to multiple cameras mounted to a vehicle, the method comprising:

• storing multiple image data captured by each of the multiple cameras;
• inputting and storing the multiple image data captured by each of the cameras into the image memory;
• managing projection surface geometry information including a set of three-dimensional coordinates for a projection surface that is virtually formed to surround the vehicle;
• determining viewing information so that a viewpoint and a viewing direction relative to the projection surface are continuously transited according to given viewing transition regulations;
• projecting and combining each of the stored image data on the projection surface, and converting the combined image data into two-dimensional image data based on the determined viewing information;
• displaying the converted two-dimensional image data;
• generating a trigger signal to make the converting start an image conversion process;
• storing the three-dimensional profile information of the vehicle;
• and
• determining and storing a minimum elevation angle for viewing so that a range of the image data shielded by the three-dimensional profile information of the vehicle has a given distance or less, when the three-dimensional profile information of the vehicle is combined with each of the image data projected on the projection surface, wherein
• the converting includes reading the stored three-dimensional profile information of the vehicle, combining the three-dimensional profile information of the vehicle with each of the image data projected to the projection surface, and converting the combined image data into two-dimensional image data, and
• the determining viewing information includes correcting the viewing information based on the stored minimum elevation angle for viewing.

An image processing method for in-vehicle image processing apparatus capable of being linked to multiple cameras mounted to a vehicle, the method comprising:

• storing multiple image data captured by each of the multiple cameras;
• inputting and storing the multiple image data captured by each of the cameras into the image memory;
• managing projection surface geometry information including a set of three-dimensional coordinates for a projection surface that is virtually formed to surround the vehicle;
• determining viewing information so that a viewpoint and a viewing direction relative to the projection surface are continuously transited according to given viewing transition regulations;
• projecting and combining each of the stored image data on the projection surface, and converting the combined image data into two-dimensional image data based on the determined viewing information;
• displaying the converted two-dimensional image data;
• generating a trigger signal to make the converting start an image conversion process;
• and
• storing the viewing transition regulations, wherein
• the stored viewing transition regulations include start-point viewing information, end-point viewing information and viewing interpolation regulations,
• the determining viewing information includes continuously changing the viewing information based on the start-point viewing information, the end-point viewing information, and the viewing interpolation regulations, and
• the stored viewing interpolation regulations stored in the viewing transition-regulation memory continuously increase a transition speed for a given period of time after the transition of viewing information is started from the start-point viewing information, and continuously decrease the transition speed for a given period of time before the transition of viewing information is stopped at the end-point viewing information.