• 検索結果がありません。

Vector Motion Low- Their Application CMOS

N/A
N/A
Protected

Academic year: 2021

シェア "Vector Motion Low- Their Application CMOS"

Copied!
1
0
0

読み込み中.... (全文を見る)

全文

(1)

Non-Destructive CMOS Image Sensors and Their Application to Low- Power Motion Vector Estimation

2001 Dwi Handoko

This thesis describes

a

new CMOS image sensors and a new block matching algorithm for on-sensor implementation of motion vector estimation in order to implement low-power video compression

system.

This work proposes an iterative block matching that

uses

high-speed intermediate pictures between adjacent video rate pictures. It estimates motion vector of video rate pictures with a reduced computational complexity while maintaining the estimation precision comparable to the full

search

block matching. Moreover, by using previously obtained motion vectors of intermediate pictures, the

search

range

can

be adaptively

extended.

In case of the search range is +30 and 8 times higher frame rate is used, the adaptive search iterative block matching, reduces the complexity of

a

full

search

block matching by

a

factor of about 1/10, while obtaining

a

comparable quality of the compressed pictures. It reduces the high-speed pictures frame rate requirement of

a

normal iterative block matching with

same search area.

Further reduction of computational complexity

can

be achieved by introducing

a

sub-sampling

search.

This work proposes also

a

CMOS image sensor suitable for the used in the iterative block matching. The image sensor captures high-speed intermediate pictures in non-destructive manner and video rate pictures with sufficient signal charge accumulation. In this

sensor,

the accumulated signal

charge

is transferred bi-directionally between two internal analog memories to obtain non-destructive high-speed pictures with

a

reduced fixed pattern noise.

The proposed sensor

has

been designed and fabricated by using 0.35pm CMOS process technology. The sensor captures 480 frame/s high-speed intermediate pictures and 30 frame/s fully accumulated video rate pictures with about 17.8mW of power dissipation

at

3.3V power supply. The proper operation of the

sensor has been

confirmed in the experiments.

An analog-to-digital conversion technique and an iterative block matching circuits for the integration to the image

sensor

chip

are

also

discussed and the

performance on low-power dissipation

have been

confirmed by circuit simulations.

‑139‑

参照

関連したドキュメント

A Tabu search procedure is then used to select a subset of financial ratio variables which best predict bankruptcy from among a larger initial set of 20 variables, and use that

Since the optimizing problem has a two-level hierarchical structure, this risk management algorithm is composed of two types of swarms that search in different levels,

Under certain assumptions on the sequence (P N ) N≥0 (which still allow for the standard probabilis- tic models of algorithms associated with binary search trees, digital search

The strategy to prove Proposition 3.4 is to apply Lemma 3.5 to the subspace X := (A p,2 ·v 0 ) ⊥ which is the orthogonal for the invariant form h·, ·i p,g of the cyclic space

We show that a discrete fixed point theorem of Eilenberg is equivalent to the restriction of the contraction principle to the class of non-Archimedean bounded metric spaces.. We

For performance comparison of PSO-based hybrid search algorithm, that is, PSO and noising-method-based local search, using proposed encoding/decoding technique with those reported

Kilbas; Conditions of the existence of a classical solution of a Cauchy type problem for the diffusion equation with the Riemann-Liouville partial derivative, Differential Equations,

Then it follows immediately from a suitable version of “Hensel’s Lemma” [cf., e.g., the argument of [4], Lemma 2.1] that S may be obtained, as the notation suggests, as the m A