IP Library › Granted Patent US 10,152,766
Granted Patent B2
US 10,152,766 · App. 14/959,724 · Granted Dec 11, 2018

Image processor, method, and chipset for increasing intergration and performance of image processing

Inventors: Seong Woon Kim (Gyeonggi-do, KR); Sung Chul Yoon (Gyeonggi-do, KR); Sang Hoon Lee (Gyeonggi-do, KR); Ha Na Yang (Gyeonggi-do, KR)
Assignee: Samsung Electronics Co., Ltd
G06T1/60G06F3/0611G06F3/0683G06T1/20G06T3/40G09G5/393G09G5/399G09G2340/00
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 10,152,766
App. No.
14/959,724
Granted
Dec 11, 2018
Kind
B2
Abstract

An image processor, an application processor, a method of operating an image processor, and a chips set of an image processor are provided. The image processor includes a scaler configured to perform scaling on an input image and generate a scaled input image; and a selection circuit configured to transmit the scaled input image to either a low latency memory or a high density memory according to a memory selection signal. The application processor includes a memory configured to store an input image; and an image processor configured to scale the input image, wherein the image processor comprises a scaler configured to perform scaling on the input image and generate a scaled input image and a selection circuit configured to transmit the scaled input image to either a low latency memory or a high density memory according to a memory selection signal.

Claims (27)

1. An image processor, comprising:

a scaler configured to perform scaling on an input image and generate a scaled input image;

a selection circuit configured to transmit the scaled input image to either a low latency memory or a high density memory according to a memory selection signal; and

a controller configured to generate the memory selection signal based on a horizontal length of the input image, a number of pixels processed per cycle of a clock signal, and a read latency of the high density memory,

wherein the memory selection signal has a level determined depending on a result of comparing a horizontal processing cycle obtained by dividing the horizontal length of the input image by the number of pixels processed per cycle of the clock signal with the read latency of the high density memory, and

wherein the scaled input image is transmitted to the high density memory when the horizontal processing cycle is greater than the read latency of the high density memory and is transmitted to the low latency memory when the horizontal processing cycle is less than or equal to the read latency of the high density memory.

2. The image processor of claim 1 , wherein the scaler performs vertical scaling on a first line of the input image at a cycle immediately following completion of horizontal scaling on the first line of the input image.

3. The image processor of claim 1 , wherein a capacity of the high density memory is determined depending on a maximum horizontal length of the scaled input image or a vertical scaling method.

4. The image processor of claim 1 , wherein a capacity of the low latency memory is determined depending on the number of pixels processed during read latency of the high density memory or a vertical scaling method.

5. The image processor of claim 1 , wherein the low latency memory is formed using at least one flip-flop.

6. An application processor, comprising:

a memory configured to store an input image;

an image processor configured to scale the input image, wherein the image processor comprises a scaler configured to perform scaling on the input image and generate a scaled input image and a selection circuit configured to transmit the scaled input image to either a low latency memory or a high density memory according to a memory selection signal; and

a controller configured to generate the memory selection signal based on a horizontal length of the input image, a number of pixels processed per cycle of a clock signal, and a read latency of the high density memory,

wherein the memory selection signal has a level determined depending on a result of comparing a horizontal processing cycle obtained by dividing the horizontal length of the input image by the number of pixels processed per cycle of the clock signal with the read latency of the high density memory, and

wherein the scaled input image is transmitted to the high density memory when the horizontal processing cycle is greater than the read latency of the high density memory and is transmitted to the low latency memory when the horizontal processing cycle is less than or equal to the read latency of the high density memory.

7. The application processor of claim 6 , wherein the scaler is further configured to perform vertical scaling on a first line of the input image at a cycle immediately following completion of horizontal scaling on the first line of the input image.

8. The application processor of claim 6 , wherein a capacity of the high density memory is determined depending on a maximum horizontal length of the scaled input image or a vertical scaling method.

9. A chipset of an image processor, the chipset configured to:

perform horizontal scaling on a first line of an input image;

transmit the first line that has been vertically scaled to a low latency memory or a high density memory according to a memory selection signal;

perform vertical scaling on the first line at a cycle following completion of the horizontal scaling on the first line; and

generate the memory selection signal based on a horizontal length of the input image, a number of pixels processed per cycle of a clock signal, and a read latency of the high density memory,

wherein the memory selection signal has a level determined depending on a result of comparing a horizontal processing cycle obtained by dividing the horizontal length of the input image by the number of pixels processed per cycle of the clock signal with the read latency of the high density memory, and

wherein a scaled input image is transmitted to the high density memory when the horizontal processing cycle is greater than the read latency of the high density memory and is transmitted to the low latency memory when the horizontal processing cycle is less than or equal to the read latency of the high density memory.

10. The chipset of an image processor of claim 9 , wherein a capacity of the high density memory is determined depending on a maximum horizontal length of the scaled input image or a vertical scaling method.

11. The chipset of an image processor of claim 9 , wherein a capacity of the low latency memory is determined depending on the number of pixels processed during read latency of the high density memory or a vertical scaling method.

Assignments (1)
ASSIGNMENT OF ASSIGNOR'S INTEREST Recorded Feb 8, 2016
From: KIM, SEONG WOON; YOON, SUNG CHUL; LEE, SANG HOON; YANG, HA NA
To: SAMSUNG ELECTRONICS CO., LTD.
Reel/Frame 037684/0951 →
Priority Claims (1)
KR 10-2014-0173133 · Dec 4, 2014 · national
Continuity (1)
Related Publication 20160163020A1 · Jun 9, 2016