Skip to content

Advertisement

  • Research
  • Open Access

Some inequalities related to ( i , j ) -type L p -mixed affine surface area and L p -mixed curvature image

Journal of Inequalities and Applications20132013:470

https://doi.org/10.1186/1029-242X-2013-470

  • Received: 29 March 2013
  • Accepted: 5 September 2013
  • Published:

Abstract

In this article, we introduce two concepts: the ( i , 0 ) -type L p -mixed affine surface area and ( i , j ) -type L p -mixed affine surface area in the set of convex bodies such that L p -affine surface area by Lutwak et al. is proposed in its special cases. Besides, applying these concepts, we establish the extension results of the well-known L p -Petty affine projection inequality, L p -Busemann centroid inequality and its dual inequality.

MSC:52A40, 52A20.

Keywords

  • convex body
  • ( i , 0 ) -type L p -mixed affine surface area
  • ( i , j ) -type L p -mixed affine surface area
  • L p -mixed curvature image

1 Introduction

Let K n denote the set of convex bodies (compact, convex subsets with nonempty interiors) in Euclidean space R n . For the set of convex bodies, which contain the origin in their interiors, and the set of origin-symmetric convex bodies in K n , we write K o n and K e n , respectively. Let S o n denote the set of star bodies (about the origin) in R n , and let S e n denote the set of origin-symmetric star bodies in S o n . Let S n 1 denote the unit sphere in R n , and let V ( K ) denote the n-dimensional volume of body K. If K is the standard unit ball B in R n , then it is denoted by ω n = V ( B ) . Note that ω n = π n / 2 Γ ( 1 + n / 2 ) defines ω n for all non-negative real n (not just the positive integers).

A body K K n is said to have a continuous i th curvature function f i ( K , ) : S n 1 [ 0 , ) if and only if S i ( K , ) is absolutely continuous with respect to S and has the Radon-Nikodym derivative (see [1])
d S i ( K , ) d S = f i ( K , ) .
(1.1)

Let F i n denote the subset of all bodies K n which have a positive continuous i th curvature function. Let F i , o n , F i , e n denote the set of all bodies in K o n , K e n , respectively, and both of them have a positive continuous i th curvature function.

A convex body K K o n is said to have an L p -curvature function f p ( K , ) : S n 1 R if its L p -surface area measure S p ( K , ) is absolutely continuous with respect to the spherical Lebesgue measure S, and it has the Radon-Nikodym derivative (see [2])
d S p ( K , ) d S = f p ( K , ) .
(1.2)
For p 1 , K F o n , then L p -affine surface area Ω p ( K ) of K by (see [35])
Ω p ( K ) = S n 1 f p ( K , u ) n n + p d S ( u ) .
(1.3)
For each K K n and p 1 , the L p -projection body, Π p K , of K is an origin-symmetric convex body whose support function is given by (see [6])
h Π p K p ( u ) = 1 n ω n c n 2 , p S n 1 | u , v | p d S p ( K , v ) , for all  u S n 1 ,
(1.4)

where c n , p = ω n + p ω 2 ω n ω p 1 . When p = 1 , (1.4) is the notion of projection body (see [7]).

It is easy to show that if E is an ellipsoid which is centered at the origin, then (see [[8], p.105])
Π p E = ( ω n V ( E ) ) 1 p E .
(1.5)

The well-known L p -Petty affine projection inequality is expressed as follows (see [810]).

Theorem A ( L p -Petty affine projection inequality)

If K , L F o n , p 1 , then
n ω n n n + p V ( Π p K ) p n + p Ω p ( K ) ,
(1.6)

with equality if and only if K is an ellipsoid which is centered at the origin.

Let K S o n , and let p 1 , then the L p -centroid body, Γ p K , of K is the origin-symmetric convex body whose support function is given by (see [6, 11])
h Γ p K p ( u ) = 1 c n , p V ( K ) K | u , x | p d x for all  u S n 1 .
(1.7)

If E is an ellipsoid which is centered at the origin, then Γ p E = E . In particular, Γ p B = B .

The well-known L p -Busemann-Petty centroid inequality is as follows (see [6]).

Theorem B ( L p -Busemann-Petty centroid inequality)

If K S o n and p 1 , then
V ( Γ p K ) V ( K ) ,
(1.8)

with the equality if and only if K is an ellipsoid which is centered at the origin.

Lutwak et al. introduced the concept of dual L p -centroid bodies (see [12]). We give the concept of the unusual normalization of dual L p -centroid bodies such that Γ p B = B : Let K K o n and real p > 0 , then radial function of dual L p -centroid body, Γ p K , of K is defined by
ρ Γ p K p ( u ) = 1 ( n + p ) c n , p V ( K ) S n 1 | u , v | p d S p ( K , v ) for all  u S n 1 .
(1.9)
It is easy to show that if E is an ellipsoid which is centered at the origin, then
Γ p E = ( V ( E ) ω n ) 1 p Π p E .
(1.10)

Combined with (1.5) and (1.10), we have that Γ p E = E . In particular, Γ p B = B .

Si Lin gives the following dual inequality of inequality (1.8) (see [[13], p.9, Theorem 5.4]).

Theorem C (Dual L p -Busemann-Petty centroid inequality)

If K K o n and p 1 , then
V ( Γ p K ) V ( K ) ,
(1.11)

with equality if and only if K is an ellipsoid which is centered at the origin.

Liu et al. [14], Lu and Wang [15], Ma and Liu [16, 17] independently proposed the notion of L p -mixed curvature function: Let p 1 , i = 0 , 1 , , n 1 , a convex body K K o n is said to have an L p -mixed curvature function f p , i ( K , ) : S n 1 R , if its L p -mixed surface area measure S p , i ( K , ) is absolutely continuous with respect to spherical Lebesgue measure S and has the Radon-Nikodym derivative
d S p , i ( K , ) d S = f p , i ( K , ) .
(1.12)
If the mixed surface area measure S i ( K , ) is absolutely continuous with respect to spherical Lebesgue measure S, we have
f p , i ( K , u ) = f i ( K , u ) h ( K , u ) 1 p .
(1.13)
According to the concept of L p -mixed curvature function of convex body, Lu and Wang [15] and Ma introduce the concept of L p -mixed curvature image of convex body as follows: For each K F i , o n ( i = 0 , 1 , , n 1 ) and real p 1 , define Λ p , i K S o n , the L p -mixed curvature image of K, by
ρ ( Λ p , i K , ) n + p i = W ˜ i ( Λ p , i K ) ω n f p , i ( K , ) .
(1.14)
If i = 0 in (1.14), then Λ p , 0 K = Λ p K . The unusual normalization of definition (1.14) is chosen so that for the unit ball B, we have Λ p , i B = B . For K F i , o n , n i p 1 , λ > 0 ,
Λ p , i λ K = λ n p i p Λ p , i K .
(1.15)
Let C i , o n denote the set of L p -mixed curvature images of convex bodies in F i , o n . That is,
C i , o n = { Q = Λ p , i L : L F i , o n } .

Because the L p -mixed curvature image belongs to star bodies, thus, C i , o n S o n .

For each K K n , real p 1 and i = 0 , 1 , , n 1 , the L p -mixed projection body, Π p , i K , of K is an origin-symmetric convex body whose support function is given by (see [18])
h Π p , i K p ( u ) = 1 n ω n c n 2 , p S n 1 | u , v | p d S p , i ( K , v ) , for all  u S n 1 ,
(1.16)
where S p , i ( K , ) ( i = 0 , 1 , , n 1 ) is L p -mixed surface area measure. Obviously, Π p , 0 K = Π p K , and for the standard unit ball B, we have Π p , i B = B . For K K n , λ > 0 , p 1 and 0 i < n , then
Π p , i ( λ K ) = λ n i p p Π p , i K .
(1.17)
Let K S o n , real p 1 and i be arbitrary real numbers, then the L p -mixed centroid body, Γ p , i K , of K is the origin-symmetric convex body whose support function is given by (see [19])
h Γ p , i K p ( u ) = 1 ( n + p ) c n , p W ˜ i ( K ) S n 1 | u , v | p ρ K n + p i ( v ) d S ( v ) for all  u S n 1 .
(1.18)

Obviously, Γ p , 0 K = Γ p K , and for the standard unit ball B, we have Γ p , i B = B .

Ma introduced the concept of dual L p -mixed centroid body (see [19]). Further, we introduced the concept of the unusual normalization of dual L p -mixed centroid body Γ p , i K as follows: Let K K o n , p > 0 , i = 0 , 1 , , n 1 , then the dual L p -mixed centroid bodies, Γ p , i K , of K are defined by:
ρ Γ p , i K p ( u ) = 1 ( n + p ) c n , p W i ( K ) S n 1 | u , v | p d S p , i ( K , v ) for all  u S n 1 .
(1.19)

Obviously, Γ p , 0 K = Γ p K , and for the standard unit ball B, we have Γ p , i B = B .

In this article, we will first introduce the concept of ( i , 0 ) -type L p -mixed affine surface area of convex body as follows.

Definition 1.1 For K F i , o n , i = 0 , 1 , , n 1 and p 1 , the ( i , 0 ) -type L p -mixed affine surface area, Ω p ( i ) ( K ) , of K is defined by
Ω p ( i ) ( K ) = S n 1 f p , i ( K , u ) n i n + p i d S ( u ) .
(1.20)

Next, we have established an extension of L p -Petty affine projection inequality (1.6) as follows.

Theorem 1.1 Let K F i , o n , i = 0 , 1 , , n 1 and p 1 , then
n ω n n i n + p i W i ( Π p , i K ) p n + p i Ω p ( i ) ( K ) ,
(1.21)

with equality in inequality (1.21) for 0 < i < n 1 if and only if K is a ball which is centered at the origin; for i = 0 if and only if K is an ellipsoid which is centered at the origin.

Further, we obtain the following generalized L p -Busemann-Petty centroid inequality.

Theorem 1.2 Suppose that K C i , o n S o n , i = 0 , 1 , , n 1 and p 1 , then
W i ( Γ p , i K ) W ˜ i ( K ) ,
(1.22)

with equality in inequality (1.22) for 0 < i < n 1 if and only if K is a ball which is centered at the origin; for i = 0 if and only if K is an ellipsoid which is centered at the origin.

Finally, we get the following dual inequality of the inequality (1.22).

Theorem 1.3 Suppose that K K o n . If Γ p , i K C i , o n S o n , i = 0 , 1 , , n 1 and p 1 , then
W ˜ i ( Γ p , i K ) W i ( K ) ,
(1.23)

with equality in inequality (1.23) for 0 < i < n 1 if and only if K is a ball which is centered at the origin; for i = 0 if and only if K is an ellipsoid which is centered at the origin.

2 Preliminaries

2.1 Support function, radial function and polar of convex body

If K K n , then its support function, h K = h ( K , ) : R n ( 0 , ) , is defined by (see [20, 21])
h ( K , x ) = max { x , y : y K } , x R n .

Obviously, if K K n , λ is a positive constant, x R n , then h ( λ K , x ) = λ h ( K , x ) .

If K is a compact star-shaped (about the origin) in R n , its radial function, ρ K = ρ ( K , ) : R n { 0 } [ 0 , ) , is defined by (see [20, 21])
ρ ( K , x ) = max { λ 0 : λ x K } , x R n { 0 } .

When ρ K is positive and continuous, K is called a star body (about the origin). Obviously, if K S o n , α > 0 , x R n , then ρ ( K , α x ) = α 1 ρ ( K , x ) and ρ ( α K , x ) = α ρ ( K , x ) . Two star bodies K and L are said to be dilates (of one another) if ρ K ( u ) / ρ L ( u ) is independent on u S n 1 .

For K K o n , the polar body, K , of K is defined by (see [20, 21])
K = { x R n : x , y 1 , y K } .
Obviously, we have ( K ) = K . If λ > 0 , then
( λ K ) = λ 1 K .
(2.1)
If K K o n , then the support and radial function of the polar body K , of K are defined respectively by (see [20, 21])
h K ( u ) = 1 ρ K ( u ) and ρ K ( u ) = 1 h K ( u )
(2.2)

for all u S n 1 .

2.2 The quermassintegrals, L p -mixed quermassintegrals and L p -mixed volume

For K K n and i = 0 , 1 , , n 1 , the quermassintegrals, W i ( K ) , of K are defined by (see [20, 21])
W i ( K ) = 1 n S n 1 h ( K , u ) d S i ( K , u ) .
(2.3)
From (2.3), we easily see that
W 0 ( K ) = V ( K ) .
(2.4)
For p 1 , K , L K o n and ε > 0 , the Firey L p -combination K + p ε L K o n is defined by (see [22])
h ( K + p ε L , ) p = h ( K , ) p + ε h ( L , ) p ,

where ‘’ in ε L denotes the Firey scalar multiplication.

For K , L K o n , ε > 0 and real p 1 , the L p -mixed quermassintegrals, W p , i ( K , L ) , of K and L ( i = 0 , 1 , , n 1 ) are defined by (see [1])
n i p W p , i ( K , L ) = lim ε 0 + W i ( K + p ε L ) W i ( K ) ε .
(2.5)
Obviously, for p = 1 , W 1 , i ( K , L ) is just the classical mixed quermassintegral W i ( K , L ) . For i = 0 , the L p -mixed quermassintegral W p , 0 ( K , L ) is just the L p -mixed volume V p ( K , L ) , namely,
W p , 0 ( K , L ) = V p ( K , L ) .
(2.6)
For p 1 , i = 0 , 1 , , n 1 and each K K o n , there exists a positive Borel measure S p , i ( K , ) on S n 1 such that the L p -mixed quermassintegral W p , i ( K , L ) has the following integral representation (see [1]):
W p , i ( K , L ) = 1 n S n 1 h L p ( v ) d S p , i ( K , v )
(2.7)
for all L K o n . It turns out that the measure S p , i ( K , ) ( i = 0 , 1 , , n 1 ) on S n 1 is absolutely continuous with respect to S i ( K , ) and has the Radon-Nikodym derivative
d S p , i ( K , ) d S i ( K , ) = h 1 p ( K , ) .
(2.8)
From (2.3) and (2.7), it follows immediately that for each K K o n ,
W p , i ( K , K ) = W i ( K ) .
(2.9)
If K , L K o n , p 1 , by definition (1.12), then formula (2.7) of the L p -mixed quermassintegral can be rewritten as follows:
W p , i ( K , L ) = 1 n S n 1 h ( L , u ) p f p , i ( K , u ) d S ( u ) .
(2.10)
We shall require the Minkowski inequality for the L p -mixed quermassintegrals W p , i as follows (see [1]): For K , L K o n , and p 1 , 0 i < n , then
W p , i ( K , L ) n i W i ( K ) n i p W i ( L ) p ,
(2.11)

with equality for p = 1 and 0 i < n 1 if and only if K and L are homothetic; for p > 1 and 0 i n 1 if and only if K and L are dilates. For p = 1 and i = n 1 , inequality (2.11) is identical.

2.3 Dual quermassintegrals and L p -dual mixed quermassintegrals

For K S o n and any real i, the dual quermassintegrals, W ˜ i ( K ) , of K are defined by (see [20, 21])
W ˜ i ( K ) = 1 n S n 1 ρ K n i ( u ) d S ( u ) .
(2.12)
Obviously,
W ˜ 0 ( K ) = V ( K ) .
(2.13)
For K , L S o n , p 1 and ε > 0 , the L p -harmonic radial combination K + p ε L S o n is defined by (see [2, 23, 24])
ρ ( K + p ε L , ) p = ρ ( K , ) p + ε ρ ( L , ) p .

Note that here ‘ ε L ’ is different from ‘ ε L ’ in the Firey L p -combination.

For K , L S o n , ε > 0 , p 1 and real i n , the L p -dual mixed quermassintegrals, W ˜ p , i ( K , L ) , of K and L are defined by [25]
n i p W ˜ p , i ( K , L ) = lim ε 0 + W ˜ i ( K + p ε L ) W ˜ i ( K ) ε .
(2.14)
If i = 0 , we easily see that definition (2.14) is just the definition of L p -dual mixed volume, namely,
W ˜ p , 0 ( K , L ) = V ˜ p ( K , L ) .
(2.15)
From definition (2.14), the integral representation of the L p -dual mixed quermassintegrals is given by (see [25]): If K , L S o n , p 1 , and real i n , i n + p , then
W ˜ p , i ( K , L ) = 1 n S n 1 ρ K n + p i ( u ) ρ L p ( u ) d S ( u ) .
(2.16)
Together with (2.12) and (2.16), for K S o n , p 1 , and i n , n + p , we get
W ˜ p , i ( K , K ) = W ˜ i ( K ) .
(2.17)
Analog of the Minkowski inequality for L p -dual mixed quermassintegrals is as follows (see [25]): If K , L S o n , p 1 , then for i < n or i > n + p ,
W ˜ p , i ( K , L ) n i W ˜ i ( K ) n + p i W ˜ i ( L ) p .
(2.18)

For n < i < n + p , inequality (2.18) is reversed. With equality in every inequality if and only if K and L are dilates.

3 The ( i , j ) -type L p -mixed affine surface area

In this section, we further propose the concept of ( i , j ) -type L p -mixed affine surface area as follows.

Definition 3.1 For K F i , o n , i = 0 , 1 , , n 1 , j R and p 1 , the ( i , j ) -type L p -mixed affine surface area, Ω p , j ( i ) ( K ) , of K is defined by
Ω p , j ( i ) ( K ) = S n 1 f p , i ( K , u ) n i j n + p i d S ( u ) .
(3.1)

Obviously, Ω p ( 0 ) ( K ) = Ω p ( K ) and Ω p , 0 ( i ) ( K ) = Ω p ( i ) ( K ) .

Next, we introduce the concept of ( i , 0 ) -type L p -mixed affine surface area of the convex bodies K 1 , K 2 , , K n i as follows.

Definition 3.2 For p 1 , i = 0 , 1 , , n 1 , the ( i , 0 ) -type L p -mixed affine surface area, Ω p ( i ) ( K 1 , , K n i ) , of K 1 , , K n i F i , o n is defined by
Ω p ( i ) ( K 1 , , K n i ) = S n 1 [ f p , i ( K 1 , u ) f p , i ( K n i , u ) ] 1 n + p i d S ( u ) .
(3.2)

Let K 1 = = K n i j = K and K n i j + 1 = = K n i = L ( j = 0 , , n i ), we define Ω p , j ( i ) ( K , L ) = Ω p ( i ) ( K , , K , L , , L ) with n i j copies of K and j copies of L. From this, if j is any real number, we can define the following.

Definition 3.3 For K , L F i , o n , i = 0 , , n 1 , p 1 , j R , the ( i , j ) -type L p -mixed affine surface area, Ω p , j ( i ) ( K , L ) , of K, L is defined by
Ω p , j ( i ) ( K , L ) = S n 1 f p , i ( K , u ) n i j n + p i f p , i ( L , u ) j n + p i d S ( u ) .
(3.3)
Specially for the case j = p , we have that
Ω p , p ( i ) ( K , L ) = S n 1 f p , i ( K , u ) f p , i ( L , u ) p n + p i d S ( u ) .
(3.4)
Take L = B in (3.4) and write
Ω p , j ( i ) ( K ) : = Ω p , j ( i ) ( K , B ) .
(3.5)
Because for u S n 1 , S i ( B , u ) = S , h ( B , u ) = 1 , so by (2.8) and (1.13), we get f p , i ( B , u ) = 1 . This together with (3.3) and (3.5) yields
Ω p , j ( i ) ( K ) = S n 1 f p , i ( K , u ) n i j n + p i d S ( u ) ,
(3.6)

and Ω p , j ( i ) ( K ) is called the ( i , j ) -type L p -mixed affine surface area of K F i , o n . In particular, Ω p , j ( o ) ( K ) = Ω p , j ( K ) is called the j th L p -mixed affine surface area of K F o n (see [26]).

Next, we give some propositions of L p -mixed curvature image and ( i , j ) -type L p -mixed affine surface area.

Proposition 3.1 Let K F i , o n , i = 0 , 1 , , n 1 , j R . Then
Ω p , j ( i ) ( K ) = n ( ω n W ˜ i ( Λ p , i K ) ) n i j n + p i W ˜ i + j ( Λ p , i K ) .
(3.7)
In particular, take j = 0 in (3.7), then
Ω p ( i ) ( K ) = n ω n n i n + p i W ˜ i ( Λ p , i K ) p n + p i .
(3.8)
Proof From (3.6), (1.14) and (2.12), we have
Ω p , j ( i ) ( K ) = S n 1 f p , i ( K , u ) n i j n + p i d S ( u ) = ( ω n W ˜ i ( Λ p , i K ) ) n i j n + p i S n 1 ρ ( Λ p , i K , u ) n i j d S ( u ) = ( ω n W ˜ i ( Λ p , i K ) ) n i j n + p i W ˜ i + j ( Λ p , i K ) .

 □

Proposition 3.2 Let p 1 , K F i , o n and i = 0 , 1 , , n 1 . Then
W p , i ( K , Q ) = ω n W ˜ i ( Λ p , i K ) W ˜ p , i ( Λ p , i K , Q )
(3.9)

for each Q K o n .

Proof For each Q K o n , from (2.10), (1.14), (2.2) and (2.16), we have
W p , i ( K , Q ) = 1 n S n 1 ρ p ( Q , u ) f p , i ( K , u ) d S ( u ) = ω n n W ˜ i ( Λ p , i K ) S n 1 ρ ( Q , u ) p ρ ( Λ p , i K , u ) n + p i d S ( u ) = ω n W ˜ i ( Λ p , i K ) W ˜ p , i ( Λ p , i K , Q ) .

 □

Proposition 3.3 If p 1 , L F i , o n , then
Ω p ( i ) ( L ) n + p i n n + p i W p , i ( L , K ) n i W ˜ i ( K ) p
(3.10)

for all K S o n with equality if and only if K and Λ p , i L are dilates.

Proof Let L F i , o n and each K S o n , then from (1.20), (2.2), (2.7), (2.12) and Hölder’s inequality, we have
Ω p ( i ) ( L ) n + p i = [ S n 1 f p , i ( L , u ) n i n + p i d S ( u ) ] n + p i = [ S n 1 ( ρ ( K , u ) p f p , i ( L , u ) ) n i n + p i ( ρ ( K , u ) n i ) p n + p i d S ( u ) ] n + p i n n + p i ( 1 n S n 1 ρ ( K , u ) p f p , i ( L , u ) d S ( u ) ) n i ( 1 n S n 1 ρ ( K , u ) n i d S ( u ) ) p = n n + p i W p , i ( L , K ) n i W ˜ i ( K ) p .

From this, we immediately get (3.10).

According to the condition of equality to hold in the Hölder inequality, we know that equality holds in (3.10) if and only if
ρ ( K , u ) p f p , i ( L , u ) ρ ( K , u ) n i = c
for any u S n 1 , where c is a constant. Combined with the definition of L p -mixed curvature image, for any u S n 1 , we have
ρ ( Λ p , i L , u ) n + p i ρ ( K , u ) n + p i = c W ˜ i ( Λ p , i L ) ω n ,

this shows that K and Λ p , i L are dilates. Therefore, the equality holds in inequality (3.10) if and only if K and Λ p , i L are dilates. The proof is complete. □

Now, according to Proposition 3.3, we can give an expansion of the definition of the ( i , 0 ) -type L p -mixed affine surface area of K K o n as follows.

Definition 3.4 If K K o n , p 1 , then the ( i , 0 ) -type L p -mixed affine surface area, Ω p ( i ) ( K ) , of K is defined by
n p n i Ω p ( i ) ( K ) n + p i n i = inf { n W p , i ( K , Q ) W ˜ i ( Q ) p n i : Q S o n } .
(3.11)

For i = 0 , the definition is just the definition of L p -affine surface area by Lutwak proposed in [2].

4 Generalized L p -Petty affine projection inequality

In this section, we complete the proof of Theorem 1.1 in the introduction. In fact, we prove the following more general conclusion.

Theorem 4.1 Let K K o n , L F i , o n , p 1 , 0 i < n 1 , then
n ω n n i n + p i W p , i ( L , Π p , i K ) n i n + p i Ω p ( i ) ( L ) W i ( K ) n i p n i + p ,
(4.1)

with equality in inequality (4.1) for 0 < i < n 1 if and only if K and L are balls of dilates which are centered at the origin; for i = 0 if and only if K and L are ellipsoids of dilates which are centered at the origin.

In order to prove the theorems above, we first give the following three lemmas.

Lemma 4.1 (See [27])

Suppose that K K o n , i R and 0 i < n , then
W ˜ i ( K ) ω n i n V ( K ) n i n ,
(4.2)

with the equality for 0 < i < n if and only if K is a ball which is centered at the origin. If i = 0 , then (4.2) is identical.

Lemma 4.2 (See [18])

Suppose that K K o n , p > 1 and 0 < i < n 1 , i is a positive integer, then
ω n i n W i ( K ) n i p p V ( Π p , i K ) n i n ω n n i p ,
(4.3)

with equality if and only if K is a ball which is centered at the origin.

Remark 4.1 The conditions of inequality (4.3) can be relaxed to p 1 and 0 i < n 1 , while the conditions of the equality that holds can be given separately. For 0 < i < n 1 and p = 1 , the inequality (4.3) is proved by Lutwak with the equality holding if and only if K is a ball (see [7]). For i = 0 and p > 1 , inequality (4.3) is proved by Lutwak et al. with the equality that holds if and only if K is an ellipsoid which is centered at the origin (see [6]). For i = 0 and p = 1 , then (4.3) is the famous Petty projection inequality (see [28]), with the equality that holds if and only if K is an ellipsoid.

Lemma 4.3 If K , L K o n , p 1 , i = 0 , 1 , , n 1 , then
W p , i ( L , Π p , i K ) = W p , i ( K , Π p , i L ) .
(4.4)

Proof From (1.16), (2.10) and the Fubini theorem, it is easy to prove Lemma 4.3. □

Proof of Theorem 4.1 For L F i , o n and any Q K o n , by inequality (3.10) and Lemma 4.1, we have
Ω p ( i ) ( L ) n + p i n n + p i ω n p i n W p , i ( L , Q ) n i V ( Q ) p ( n i ) n ,
(4.5)

with equality for 0 < i n 1 if and only if Λ p , i L and Q are centered balls of dilates; for i = 0 if and only if Λ p , i L and Q are dilates.

Take Q = Π p , i K with K K o n in (4.5), then
Ω p ( i ) ( L ) n + p i n n + p i ω n p i n W p , i ( L , Π p , i K ) n i V ( Π p , i K ) p ( n i ) n ,
(4.6)

with equality for 0 < i n 1 if and only if Λ p , i L and Π p , i K are centered balls of dilates; for i = 0 if and only if Λ p L and Π p K are dilates.

Combining with inequalities (4.3) and (4.6), we give
Ω p ( i ) ( L ) n + p i W i ( K ) n i p n n + p i ω n p i n W p , i ( L , Π p , i K ) n i W i ( K ) n i p V ( Π p , i K ) p ( n i ) n n n + p i W p , i ( L , Π p , i K ) n i ω n n i ,

which implies that inequality (4.1) holds.

Next, we discuss the conditions of equality that holds in inequality (4.1).

According to the condition of the equality that holds in inequality (4.3) and inequality (4.6) with Remark 4.1, the four steps will be given.
  1. (1)

    For the case p > 1 and 0 < i < n 1 , the equality holds in (4.1) if and only if Π p , i K and Λ p , i L are balls of dilates which are centered at the origin, and K is a ball which is centered at the origin. Together with Λ p , i B = B and Π p , i B = B , we know that K and L are balls of dilates which are centered at the origin.

     
  2. (2)

    For the case p = 1 and 0 < i < n 1 , the equality holds in (4.1) if and only if Π 1 , i K and Λ 1 , i L are balls of dilates which are centered at the origin, and K is a ball. By using Π p , i B = B and (1.17), it is obtained that Π 1 , i ( λ B ) = λ 1 + i n Π 1 , p B = λ 1 + i n B ( λ > 0 ) is a ball which is centered at the origin. Because Π 1 , i K and Λ 1 , i L are balls of dilates which are centered at the origin, then Λ 1 , i L is a ball of dilates which are centered at the origin, and together with Λ p , i B = B and (1.15), L is a ball which is centered at the origin. However, K is a ball, so the equality holds in (4.1) if and only if K and L are balls of dilates which are centered at the origin.

     
  3. (3)

    For the case p > 1 and i = 0 , the equality holds in (4.1) if and only if Π p K and Λ p L are dilates and K is an ellipsoid which is centered at the origin. Let K = E be an ellipsoid which is centered at the origin, from (1.5), we know that Π p E = ( ω n / V ( E ) ) 1 p E is an ellipsoid which is centered at the origin. Other, from the literature [2], we know that L is an ellipsoid E which is centered at the origin if and only if Λ p L are dilates of polar body E of this E. So we know that the equality holds in (4.1) if and only if L and K are ellipsoids which are centered at the origin and both are dilates.

     
  4. (4)

    For the case p = 1 and i = 0 , the equality holds in (4.1) if and only if Λ 1 L and Π 1 K are dilates, and K is an ellipsoid. Suppose that K = λ E + x 0 with λ > 0 , x 0 R n , and E is an ellipsoid which is centered at the origin, noting that S ( λ E + x 0 , ) = S ( λ E , ) = λ n 1 S ( E , ) (see [29]), this together with (1.5) Π 1 K = Π 1 ( λ E + x 0 ) = Π 1 λ E = λ 1 n Π 1 E = λ 1 n ( ω n / V ( E ) ) E is an ellipsoid which is centered at the origin. Because Λ 1 L and Π 1 K are dilates, then Λ 1 L is an ellipsoid which is centered at the origin. However, from [2], we know that L is an ellipsoid E which is centered at the origin if and only if Λ 1 L are dilates of polar body E of this E. Therefore, the equality holds in (4.1) if and only if K and L are ellipsoids of the dilates which are centered at the origin.

     

To sum up, the equality holds in (4.1) for p 1 and 0 < i < n 1 if and only if K and L are balls of the dilates which are centered at the origin; for p 1 and i = 0 if and only if K and L are ellipsoids of the dilates which are centered at the origin. The proof is complete. □

Proof of Theorem 1.1 Exchange K and L in inequality (4.1), we have for L K o n , K F i , o n , p 1 , 0 i < n 1
n ω n n i n + p i W p , i ( K , Π p , i L ) n i n + p i Ω p ( i ) ( K ) W i ( L ) n i p n i + p .
(4.7)
By using Lemmas 4.3 and (4.7), we have
n ω n n i n + p i W p , i ( L , Π p , i K ) n i n + p i = n ω n n i n + p i W p , i ( K , Π p , i L ) n i n + p i Ω p ( i ) ( K ) W i ( L ) n i p n i + p .

Taking L = Π p , i K in the inequality above, we immediately obtain inequality (1.21). The proof is complete. □

Combining with Theorem 1.1 and (3.8), we immediately obtain the following.

Corollary 4.1 If K F i , o n , i = 0 , 1 , , n 1 and p 1 , then
W i ( Π p , i K ) W ˜ i ( Λ p , i K ) ,
(4.8)

with the equality in inequality (4.8) for 0 < i < n 1 if and only if K is a ball which is centered at the origin; for i = 0 if and only if K is an ellipsoid which is centered at the origin.

Further, we have established the following results.

Theorem 4.2 Let K , L F i , o n , 0 i < n 1 , p 1 , then
n ω n n i n + p i W p , i ( Π p , i K , Π p , i L ) p n + p i Ω p ( i ) ( K ) n p i n i Ω p ( i ) ( L ) p n i ,
(4.9)

with the equality in inequality (4.9) for 0 < i < n 1 if and only if K and L are balls of dilates which are centered at the origin; for i = 0 if and only if K and L are ellipsoids of dilates which are centered at the origin.

Proof From inequality (4.1), we know that for Q K o n , L F i , o n , p 1 , 0 i n 1 ,
n ω n n i n + p i W p , i ( L , Π p , i Q ) n i n + p i Ω p ( i ) ( L ) W i ( Q ) n i p n i + p .
(4.10)
Take Q = Π p , i K in (4.9), and using Corollary 4.1 and Lemma 4.3, we have
Ω p ( i ) ( L ) W ˜ i ( Λ p , i K ) n i p n i + p Ω p ( i ) ( L ) W i ( Π p , i K ) n i p n i + p n ω n n i n + p i W p , i ( L , Π p , i Π p , i K ) n i n + p i = n ω n n i n + p i W p , i ( Π p , i K , Π p , i L ) n i n + p i .
(4.11)

Combining inequality (4.11) with (3.8), we immediately obtain inequality (4.8). According to the condition of the equality holding in inequalities (4.1) and (4.9), the condition of the equality that holds in inequality (4.8) is easily obtained. The proof is complete. □

5 Generalized L p -Busemann-Petty centroid inequality and dual inequality

In this section, we give the extension of the well-known L p -Busemann-Petty centroid inequality (1.8). Namely, we complete the proof of Theorem 1.2 and Theorem 1.3 (i.e., dual inequality of Theorem 1.2) in the introduction.

Lemma 5.1 If K F i , o n , 0 = 0 , 1 , , n 1 , p 1 , then
Π p , i K = Γ p , i Λ p , i K .
(5.1)

Proof Using definition (1.16) of L p -mixed projection body and definition (1.14) of L p -mixed curvature image, it is easy to prove (5.1). □

Lemma 5.2 If K K o n , L S o n , p 1 , i = 0 , 1 , , n 1 , then
W p , i ( K , Γ p , i L ) W i ( K ) = W ˜ p , i ( L , Γ p , i K ) W ˜ i ( L ) .
(5.2)

Proof By (1.18), (1.19), (2.3), (2.7), (2.12) and (2.16), it is easy to prove (5.2). □

Proof of Theorem 1.2 For L F i , o n , using (5.1) and Corollary 4.1, we have
W i ( Γ p , i Λ p , i L ) = W i ( Π p , i L ) W ˜ i ( Λ p , i L ) ,

taking K = Λ p , i L in the inequality above, we immediately get inequality (1.22).

According to the condition of the equality that holds in inequalities (4.8) and (1.8), and noting that Λ p , i B = B and (1.15), we know with the equality in inequality (1.22) for p 1 and 0 < i < n 1 if and only if K is a ball which is centered at the origin; for p 1 and i = 0 if and only if K is an ellipsoid which is centered at the origin. □

Proof of Theorem 1.3 Take L = Γ p , i K in (5.2), we have
W i ( K ) = W p , i ( K , Γ p , i Γ p , i K ) .
Using the Minkowski inequality (2.11) of the L p -mixed quermassintegrals, we have
W i ( K ) W i ( K ) n i p n i W i ( Γ p , i Γ p , i K ) p n i .
(5.3)
Together with inequality (1.22), we can get
W i ( K ) W i ( Γ p , i Γ p , i K ) W ˜ i ( Γ p , i K ) ,

from this, we can get inequality (1.23).

According to the condition of the equality that holds in inequalities (2.11) and (1.22), we discuss the conditions of the equality that holds in (1.23) in the following four cases:
  1. (1)

    For the case p > 1 and 0 < i < n 1 , the equality holds in (1.23) if and only if K and Γ p , i Γ p , i K are dilates, and Γ p , i K is a ball which is centered at the origin. Because Γ p , i B = B , then Γ p , i K is a ball which is centered at the origin if and only if K is a ball which is centered at the origin. Therefore, K is a ball which is centered at the origin. While the equality Γ p , i B = B shows that Γ p , i K is a ball which is centered at the origin if and only if Γ p , i Γ p , i K is a ball which is centered at the origin. From this, for p > 1 and 0 < i < n 1 the equality holds in inequality (1.23) if and only if K is a ball which is centered at the origin.

     
  2. (2)

    For the case p > 1 and i = 0 , the equality holds in (1.23) if and only if K and Γ p Γ p K are dilates and Γ p K is an ellipsoid which is centered at the origin. Because Γ p K is an ellipsoid which is centered at the origin, and together with (1.10), K is an origin-symmetric ellipsoid E if and only if Γ p K is an origin-symmetric ellipsoid. On the other hand, the literature [11] tells us that if E is an ellipsoid which is centered at the origin, then Γ p E = E . From this, Γ p Γ p K is an ellipsoid which is centered at the origin. Therefore, for p > 1 and i = 0 , the equality holds in inequality (1.23) if and only if K is an ellipsoid which is centered at the origin.

     
  3. (3)

    For the case p = 1 and 0 < i < n 1 , the equality holds in (1.23) if and only if K and Γ 1 , i Γ 1 , i K are homothetic, and Γ 1 , i K is a ball which is centered at the origin. From Γ p , i B = B , we know that Γ 1 , i B = B , then K is a ball which is centered at the origin. This Γ 1 , i B = B together with Γ 1 , i B = B , then Γ 1 , i Γ 1 , i K is a ball which is centered at the origin. Therefore, for p = 1 and 0 < i < n 1 , the equality holds in inequality (1.23) if and only if K is a ball which is centered at the origin.

     
  4. (4)

    For the case p = 1 and i = 0 , the equality holds in (1.23) if K and Γ Γ 1 K are homothetic, and Γ 1 K is an ellipsoid which is centered at the origin. Because Γ 1 K is an origin-symmetric ellipsoid E if and only if K is an origin-symmetric ellipsoid. On the other hand, from Γ p E = E , we know that Γ 1 K is an ellipsoid which is centered at the origin if and only if Γ 1 Γ 1 K is an ellipsoid which is centered at the origin. Therefore, for p = 1 and i = 0 , the equality holds in inequality (1.23) if and only if K is an ellipsoid which is centered at the origin.

     

To sum up, the equality holds in (1.23) for p 1 and 0 < i < n 1 if and only if K is a ball which is centered at the origin; for p 1 and i = 0 if and only if K is an ellipsoid which is centered at the origin. The proof is complete. □

Declarations

Acknowledgements

The author is supported by the NNSF of China (11161019, 11371224).

Authors’ Affiliations

(1)
School of Mathematics and Statistics, Hexi University, Zhangye, Gansu, 734000, People’s Republic of China

References

  1. Lutwak E: The Brunn-Minkowski-Firey theory I: mixed volumes and the Minkowski problem. J. Differ. Geom. 1993, 38: 131–150.MathSciNetGoogle Scholar
  2. Lutwak E: The Brunn-Minkowski-Firey theory II: affine and geominimal surface areas. Adv. Math. 1996, 118: 244–294. 10.1006/aima.1996.0022MathSciNetView ArticleGoogle Scholar
  3. Lutwak E: Mixed affine surface area. J. Math. Anal. Appl. 1987, 125: 351–360. 10.1016/0022-247X(87)90097-7MathSciNetView ArticleGoogle Scholar
  4. Leichtweiß K: Bemerkungen zur definition einer erweiterten affinoberfläche von E. Lutwak. Manuscr. Math. 1989, 65: 181–197. 10.1007/BF01168298View ArticleGoogle Scholar
  5. Leichtweiß K: Affine Geometry of Convex Bodies. J. A. Barth, Heidelberg; 1998.Google Scholar
  6. Lutwak E, Yang D, Zhang GY: L p -Affine isoperimetric inequalities. J. Differ. Geom. 2000, 56: 111–132.MathSciNetGoogle Scholar
  7. Lutwak E: Mixed projection inequalities. Trans. Am. Math. Soc. 1985, 287: 91–106. 10.1090/S0002-9947-1985-0766208-7MathSciNetView ArticleGoogle Scholar
  8. Wang, WD: Extremum problems in convex bodies geometry of L p -space. A dissertation submitted to Shanghai University for the degree of doctor in science 2, 113–115 (2006) (in Chinese)Google Scholar
  9. Lü SJ, Leng GS: The L p -curvature images of convex bodies and L p -projection bodies. Proc. Indian Acad. Sci. Math. Sci. 2008, 118(3):413–424. 10.1007/s12044-008-0032-6MathSciNetView ArticleGoogle Scholar
  10. Yuan J, Lü SJ, Leng GS: The p -affine surface area. Math. Inequal. Appl. 2007, 10(3):693–702.MathSciNetGoogle Scholar
  11. Lutwak E, Zhang GY: Blaschke-Santal’o inequalities. J. Differ. Geom. 1997, 47: 1–16.MathSciNetGoogle Scholar
  12. Lutwak E, Yang D, Zhang GY: L p -John ellipsoids. Proc. Lond. Math. Soc. 2005, 90: 497–520. 10.1112/S0024611504014996MathSciNetView ArticleGoogle Scholar
  13. Lin, S: Extremum problems in convex and discrete geometry. A dissertation submitted to Shanghai University for the degree of doctor in science (2006) (in Chinese)Google Scholar
  14. Liu LJ, Wang W, He BW: Fourier transform and L p -mixed projection bodies. Bull. Korean Math. Soc. 2010, 47: 1011–1023. 10.4134/BKMS.2010.47.5.1011MathSciNetView ArticleGoogle Scholar
  15. Lu FH, Wang WD: Inequalities for L p -mixed curvature images. Acta Math. Sci. 2010, 30: 1044–1052. (in Chinese) 10.1016/S0252-9602(10)60101-4View ArticleGoogle Scholar
  16. Ma TY, Liu CY: The generalized Busemann-Petty problem for dual L p -mixed centroid bodies. J. Southwest Univ., Nat. Sci. Ed. 2012, 34(4):105–112. (in Chinese)Google Scholar
  17. Ma TY, Liu CY: The generalized Shephard problem for L p -mixed projection bodies and Minkowski-Funk transforms. J. Shandong Univ. (Nat. Sci.) 2012, 47(10):21–30. (in Chinese)MathSciNetGoogle Scholar
  18. Wang WD, Leng GS: The Petty projection inequality for L p -mixed projection bodies. Acta Math. Sin. Engl. Ser. 2007, 23(8):1485–1494. (in Chinese) 10.1007/s10114-005-0894-9MathSciNetView ArticleGoogle Scholar
  19. Ma TY: On L p -mixed centroid bodies and dual L p -mixed centroid bodies. Acta Math. Sin., Chin. Ser. 2010, 53(2):301–314. (in Chinese)Google Scholar
  20. Gardner RJ: Geometric Tomography. Cambridge University Press, Cambridge; 1995.Google Scholar
  21. Schneider R: Convex Bodies: The Brunn-Minkowski Theory. Cambridge University Press, Cambridge; 1993.View ArticleGoogle Scholar
  22. Firey WJ: p -Means of convex bodies. Math. Scand. 1962, 10: 17–24.MathSciNetGoogle Scholar
  23. Firey WJ: Polar means of convex bodies and a dual to the Brunn-Minkowski theorem. Can. J. Math. 1961, 13: 444–453. 10.4153/CJM-1961-037-0MathSciNetView ArticleGoogle Scholar
  24. Firey WJ: Mean cross-section measures of harmonic means of convex bodies. Pac. J. Math. 1961, 11: 1263–1266. 10.2140/pjm.1961.11.1263MathSciNetView ArticleGoogle Scholar
  25. Wang WD, Leng GS: L p -Dual mixed quermassintegrals. Indian J. Pure Appl. Math. 2005, 36(4):177–188.MathSciNetGoogle Scholar
  26. Wang WD, Leng GS: L p -Mixed affine surface area. J. Math. Anal. Appl. 2007, 54: 1–14.MathSciNetView ArticleGoogle Scholar
  27. Lutwak E: Dual mixed volumes. Pac. J. Math. 1975, 58(2):531–538. 10.2140/pjm.1975.58.531MathSciNetView ArticleGoogle Scholar
  28. Petty CM: Isoperimetric problems. In Proc.Conf. Convexity and Combinatorial Geometry. University of Oklahoma, Oklahoma; 1972:26–41. Univ. Oklahoma, 1971Google Scholar
  29. Wang WD, Wei DJ, Xiang Y: Some inequalities for the L p -curvature image. J. Inequal. Appl. 2009., 2009: Article ID 320786 10.1155/2009/320786Google Scholar

Copyright

© Ma; licensee Springer. 2013

This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Advertisement