JPS6217873A - Image processing system - Google Patents

Abstract

PURPOSE:To execute access from a host to a buffer and processing in an image processing part in parallel and to execute image processing at a high speed by forming an input buffer and an output buffer as a double buffer. CONSTITUTION:In the image processing by a terminal device or the like, data to be processed are written in a buffer IB0 from the host H at a time t0. At a time t1 from the end of writing, the contents of the buffer IB0 are read out, and while being processed by the image processing part IP, the processed data are written in a buffer OB0. At a time t2, data to be processed are written from the host H in a buffer IB1. At the time t2 after the end of the writing, the contents of the buffer IB1 are read out, and while being processed by the image processing part IP, the processed data are written in a buffer OB1. Simultaneously, the processed data are read out from the buffer OB0 by the host H. At a time t4, the data to be processed are written in the buffer IB0 from the host H. Thereafter, the similar operation is repeated until the data to be processed completely removed.

Description

[Detailed description of the invention] [Industrial application field] The present invention relates to image processing in terminal devices and the like.

In particular, it relates to high-speed control methods.

[Conventional technology]

Conventionally, this type of image processing has been performed using an input buffer IB between a host (host) H and an image processing unit IP, as shown in FIG.
(hereinafter abbreviated as buffer IB) and output buffer OB (hereinafter abbreviated as buffer OB).

The control is to write the data to be processed from the host H to the buffer IB, and after the writing is completed, read the buffer IB, process it in the image processing unit IP, and write it to the buffer OB.After the writing is completed, the processed data is written from the buffer OB by the host It was a serial read control.

[Problem that the invention seeks to solve]

In the conventional image processing described above, the buffer IB is
, processing in the 2nd image processing unit IP is temporarily interrupted while accessing the buffer OB, and conversely, access to the buffers IB and -4 software OB from the host H being processed in the 2nd image processing unit IP is prohibited. The drawback was that processing could only be done serially.

The present invention seeks to eliminate these drawbacks.

[Failure to solve the problem]

The image processing method of the present invention includes buffer IB and buffer O.
A feature is that each of B has a duplex configuration.

〔Example〕

Next, two embodiments of the present invention will be described with reference to the drawings.

FIG. 1 is a block diagram of one embodiment of the present invention.

The image processing method of the present invention uses buffers IBO and rB1 and buffers OBO with a double-gun configuration.

It is composed of OBI and an image processing unit IP.

FIG. 2 is a processing time chart for explaining the control operation of this method. Time t. (H) Write the data to be processed to Gunofa IBO from H. t1 after writing is completed
At the same time as reading the buffer IBO, the 5 image processing unit IP
Write to buffer OBO while processing. At time t2, the data to be processed is written from the host (H) to the buffer TBI.

After writing is completed, at t3, the buffer IB1 is read out, and while processing is performed by the second image processing unit IP, the buffer OE is read out.
Write to 1. At the same time, the data after the host H data processing is read from the pathson A OBO. At time t4, the data to be processed is again written from the host (H) to the buffer IBO. The process is repeated in the same manner until there is no more data to be processed.

〔Effect of the invention〕

As explained above, in the present invention, the input panel TB and the output panel Z'OB between the post IH and the image processing unit IP are made into a pull buffer. , access to buffer IB and buffer OR from host L-H and processing at image processing unit IP.
3° has the effect of being able to perform 9 image processing at high speed since it can be performed in 9 parallels and processing can be performed continuously in one image processing section.

[Brief explanation of drawings]

FIG. 1 is a block diagram of one embodiment of the present invention, and FIG. 2 is a processing tights chart 1 for explaining the control operation of this system.
FIG. 3 is a block diagram of a conventional image processing method. In the figure, H is a host l-, IP is an image processing unit, and TB. IBo, 1131: Inzonotopanopha, OB. OBO, OBI: Outpnotopanofa.

Claims (1)

[Claims] 1. In image processing in a terminal device or the like, an image processing method characterized in that an input buffer and an output buffer between a host and an image processing section each have a duplex configuration.